From db4109603721fb0d9edb22e4733a62a0b5615b40 Mon Sep 17 00:00:00 2001
From: chris waisi <kingwaisi@gmail.com>
Date: Sun, 7 Jul 2024 17:24:54 +0200
Subject: [PATCH] updatenewtags1
---
...{notebook (4).ipynb => notebook (5).ipynb} | 1457 +++++++--
.../notebook.ipynb | 2641 -----------------
2 files changed, 1195 insertions(+), 2903 deletions(-)
rename Tourism/Prediction cancellation of hotel bookings/{notebook (4).ipynb => notebook (5).ipynb} (98%)
delete mode 100644 Tourism/Prediction cancellation of hotel bookings/notebook.ipynb
diff --git a/Tourism/Prediction cancellation of hotel bookings/notebook (4).ipynb b/Tourism/Prediction cancellation of hotel bookings/notebook (5).ipynb
similarity index 98%
rename from Tourism/Prediction cancellation of hotel bookings/notebook (4).ipynb
rename to Tourism/Prediction cancellation of hotel bookings/notebook (5).ipynb
index 2d33178..c36522a 100644
--- a/Tourism/Prediction cancellation of hotel bookings/notebook (4).ipynb
+++ b/Tourism/Prediction cancellation of hotel bookings/notebook (5).ipynb
@@ -3,99 +3,92 @@
{
"cell_type": "markdown",
"metadata": {
+ "Paragraph": "Analyse der Bewegung und Aktivität freilaufender Rinder",
"editable": true,
"include": true,
- "paragraph": "Analyse der Bewegung und Aktivität von freilaufenden Rindern",
"slideshow": {
"slide_type": ""
},
"tags": []
},
"source": [
- "# Analyse der Bewegung und Aktivität von freilaufenden Rindern"
+ "# Analyse der Bewegung und Aktivität freilaufender Rinder"
]
},
{
"cell_type": "markdown",
"metadata": {
+ "Paragraph": "Geschäftsverstandnis",
"editable": true,
- "include": false,
- "paragraph": "Geschäftsverstandnis",
+ "include": true,
"slideshow": {
"slide_type": ""
},
"tags": []
},
"source": [
- "# 1. Geschäftsverstandnis"
+ "# 1. Geschäftsverstandinis\n",
+ "Landwirte geraten immer weiter unter wirtschaftlichen Druck. Beispielsweise kam es durch die Coronapandemie zu teuren logistischen Problemen. Gleichzeitig stellen immer mehr Konsumenten höhere Anforderungen an eine artgerechte Tierhaltung. Diese Marktbedingungen erscheinen für Landwirte problematisch, da letzterer Aspekt natürlich mit gewissen Kosten verbunden ist und ein generelles Umdenken notwendig macht. Durch neue Technologien wie z.B. Künstliche Intelligenz und der Verwendung von Sensoren können die Prozesse effizienter gestaltet und gleichzeitig die Anforderungen des Tierschutzes besser erfüllt werden. Ein Problem der Landwirte ist, dass sie ihre Tiere nicht gezielt bzw. individuell umsorgen können. Das Wissen darüber, welche Aktivitäten z.B. besonderen Einfluss auf die Milchproduktion oder Fleischqualität haben, kann bisher nicht umfänglich erfasst werden. Somit sind auch keine gezielten Maßnahmen zur Förderung bestimmter Aktivitäten möglich. Denkbare Maßnahmen wären die gezielte Motivation zu mehr Bewegung einzelner Tiere oder die Optimierung der Zucht. Bisher ist es jedoch schwierig kontinuierliche Aktivitätsprofile der Tiere zu erstellen. Ein weiteres Problem der Landwirte ist, dass eine frühzeitige Erkennung von Veränderungen des Gesundheitszustandes einzelner Tiere oft nicht möglich ist. Zwar ist die regelmäßige Kontrolle des Zustandes einer Herde rechtlich vorgeschrieben, jedoch können gerade in größeren Betrieben auf Grund der großen Anzahl an Tieren, gesundheitliche Veränderungen erst spät erkannt werden. Oft bleibt im Krankheitsfall nur die Behandlung mit Antibiotika oder anderen teuren Medikamenten. Gewissen Verhaltensänderungen können jedoch frühzeitig erste Hinweise liefern. Gesunde Rinder beispielsweise bewegen sich normalerweise recht viel und reduzieren ihre Aktivität, wenn sie erste Krankheitszeichen entwickeln. Eine frühe Erkennung dieser Bewegungsreduzierung wäre von großer Bedeutung für einen Landwirt. Healthy Cattle plant, zur Lösung der oben genannten Probleme, eine Aktivitätsklassifizierung zur Erstellung von Aktivitätsprofilen und zur frühzeitigen Erkennung von Krankheiten zu entwickeln. Die Hypothese ist, dass mit Hilfe von typischen Sensoren eines Smartphones und der Verwendung eines Machine Learning Algorithmus, die Aktivitäten klassifiziert werden können. Healthy Cattle möchte für die geplante Lösung zunächst testen, ob mit Hilfe von typischen Sensoren eines Smartphones eine ausreichend genaue Klassifizierung von Aktivitäten möglich ist. Um dies möglichst kostengünstig testen zu können, werden für einen ersten Prototypen keine eigenen Daten verwendet, sondern im Internet zugängliche Daten aus einem ähnlichen Setup genutzt"
]
},
{
"cell_type": "markdown",
"metadata": {
- "include": true,
- "paragraph": "",
+ "Include": true,
+ "Paragraph": "Datenverständnis",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
"tags": []
},
"source": [
- "Landwirte geraten immer weiter unter wirtschaftlichen Druck. Gleichzeitig\n",
- "stellen immer mehr Konsumenten höhere Anforderungen an eine artgerechte\n",
- "Tierhaltung. Diese Marktbedingungen erscheinen für Landwirte problematisch,\n",
- "da letzterer Aspekt natürlich mit gewissen Kosten verbunden ist und ein\n",
- "generelles Umdenken notwendig macht.\n",
- "\n",
- "Durch neue Technologien wie Künstliche Intelligenz und die Verwendung\n",
- "von Sensoren können Prozesse effizienter gestaltet und gleichzeitig die\n",
- "Anforderungen des Tierschutzes besser erfüllt werden.\n",
- "\n",
- "Ein Problem der Landwirte ist, dass sie ihre Tiere nicht gezielt bzw. individuell\n",
- "umsorgen können. Das Wissen darüber, welche Aktivitäten besonderen\n",
- "Einfluss auf die Milchproduktion oder Fleischqualität haben, kann bisher nicht\n",
- "umfänglich erfasst werden. Somit sind auch keine gezielten Maßnahmen zur\n",
- "Förderung bestimmter Aktivitäten möglich. Denkbare Maßnahmen wären die\n",
- "gezielte Motivation zu mehr Bewegung einzelner Tiere oder die Optimierung der\n",
- "Zucht. Bisher ist es jedoch schwierig kontinuierliche Aktivitätsprofile der Tiere zu erstellen.\n",
- "\n",
- "Ein weiteres Problem der Landwirte ist, dass eine frühzeitige Erkennung von\n",
- "Veränderungen des Gesundheitszustandes einzelner Tiere oft nicht möglich ist.\n",
- "Zwar ist die regelmäßige Kontrolle des Zustandes einer Herde rechtlich\n",
- "vorgeschrieben, jedoch können gerade in größeren Betrieben auf Grund der\n",
- "großen Anzahl an Tieren, gesundheitliche Veränderungen erst spät erkannt\n",
- "werden. \n",
- "\n",
- "Verhaltensänderungen können jedoch frühzeitig erste Hinweise liefern. Gesunde Rinder beispielsweise bewegen sich normalerweise recht viel und reduzieren ihre Aktivität, wenn sie erste Krankheitszeichen entwickeln. Eine frühe Erkennung dieser Bewegungsreduzierung wäre von großer Bedeutung für einen Landwirt."
+ "# 2. Datenverständnis\n",
+ "Der Datensatz dieser Hausarbeit stammt von Anguita, Ghio et al. und wurde abgerufen von Kaggle.com unter: https://www.kaggle.com/uciml/human-activity-recognition-with-smartphones. Der Original Datensatz ist auf https://archive.ics.uci.edu/ml/datasets/human+activity+recognition+using+smartphones. In einem Experiment haben 30 Probanden mit einem auf Hüfthöhe befestigten Smartphone einen festgelegten Bewegungsablauf durchgeführt. Dabei wurden mit Hilfe der verbauten Sensoren die linearen Beschleunigungs- und Winkelgeschwindigkeitssignale auf jeweils drei Achsen erfasst. Die einzelnen Datenpunkte wurden je nach Aktivität klassifiziert bzw. gelabelt. Durch verschiedene Filterverfahren wurden diese anschließend von Rauschen befreit und weitere zeit- der frequenzbezogene Features extrahiert. Aus diesen wurden weitere Features wie z.B. Mittelwert, Standardabweichung, Minimal- und Maximalwert, Median usw. berechnet. Das zu Grunde gelegte Machine Learning Problem ist in den Bereich Supervised Learning einzuordnen, da auf gelabelten Daten gelernt wird. Jeder Datenpunkt im Datensatz besitzt eine Reihe von Features - berechnet aus der Sensorwerten - und ein Target - die Aktivitätsklasse. Der zur Verfügung stehende Datensatz besteht aus 10299 Beobachtungen (Observations). Es gibt keine fehlenden Werte und es existieren auch keine Duplikate. Die Anzahl der Features ist sehr groß, was eine umfängliche Analyse der einzelnen Features stark erschwert. Eine grafische Überprüfung auf Multikollinearität mit Hilfe der Korrelationsmatrix ist deshalb nicht sinnvoll. Eine Überprüfung des Variance Inflation Factor (VIF) ergibt, dass sehr viele Features mit einem VIF>10 eine hohe Multikollinearität aufweisen (siehe Tabelle). Eine hohe Multikollinearität hat negative Auswirkungen auf die Interpretierbarkeit der später erstellten Modelle\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
- "include": true,
- "paragraph": "Daten und Datenverständnis"
+ "Include": true,
+ "Paragraph": "Import von Relevant Module",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
},
"outputs": [],
"source": [
- "# 2. Daten und Datenverständnis"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "include": true,
- "paragraph": "Import von relevanten Module"
- },
- "source": [
- "\n",
- "## 2.1. Import von relevanten Daten "
+ "## 2.1. Import von Relevant Modules "
]
},
{
"cell_type": "code",
- "execution_count": 34,
+ "execution_count": 2,
"metadata": {
- "include": true
+ "Include": false,
+ "Paragraph": "Import von Relevant Module",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
},
- "outputs": [],
+ "outputs": [
+ {
+ "ename": "ModuleNotFoundError",
+ "evalue": "No module named 'plotly'",
+ "output_type": "error",
+ "traceback": [
+ "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+ "\u001b[1;31mModuleNotFoundError\u001b[0m Traceback (most recent call last)",
+ "Cell \u001b[1;32mIn[2], line 4\u001b[0m\n\u001b[0;32m 1\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mnumpy\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m \u001b[38;5;21;01mnp\u001b[39;00m \u001b[38;5;66;03m# standard for data processing\u001b[39;00m\n\u001b[0;32m 2\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mpandas\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m \u001b[38;5;21;01mpd\u001b[39;00m \u001b[38;5;66;03m# standard for data processing\u001b[39;00m\n\u001b[1;32m----> 4\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mplotly\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mgraph_objects\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m \u001b[38;5;21;01mgo\u001b[39;00m \u001b[38;5;66;03m# creates plots\u001b[39;00m\n\u001b[0;32m 5\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mseaborn\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m \u001b[38;5;21;01msns\u001b[39;00m\n\u001b[0;32m 6\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01mmatplotlib\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m pyplot \u001b[38;5;28;01mas\u001b[39;00m plt\n",
+ "\u001b[1;31mModuleNotFoundError\u001b[0m: No module named 'plotly'"
+ ]
+ }
+ ],
"source": [
"import numpy as np # standard for data processing\n",
"import pandas as pd # standard for data processing\n",
@@ -126,8 +119,13 @@
{
"cell_type": "markdown",
"metadata": {
- "include": "true",
- "paragraph": "Daten Auslesen"
+ "editable": true,
+ "include": true,
+ "paragragph": "Daten Auslesen",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
},
"source": [
"## Daten Auslesen"
@@ -136,9 +134,9 @@
{
"cell_type": "markdown",
"metadata": {
+ "Include": true,
+ "Paragraph": "Daten Auslesen",
"editable": true,
- "include": true,
- "paragraph": "DataUnderstanding",
"slideshow": {
"slide_type": ""
},
@@ -147,29 +145,37 @@
"source": [
"Quelle der Daten:\n",
"\n",
- "Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra und Jorge L. Reyes-Ortiz. Ein öffentlich zugänglicher Datensatz für die Erkennung menschlicher Aktivitäten mit Smartphones. 21. Europäisches Symposium über Künstliche Neuronale Netze, Computational Intelligence und Maschinelles Lernen, ESANN 2013. Brügge, Belgien 24-26 April 2013.\n",
- "Heruntergeladen von Kaggle: https://www.kaggle.com/uciml/human-activity-recognition-with-smartphones\n"
+ "Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra und Jorge L. Reyes-Ortiz. Ein gemeinfreier Datensatz zur Erkennung menschlicher Aktivitäten mithilfe von Smartphones. 21. Europäisches Symposium zu künstlichen neuronalen Netzen, Computerintelligenz und maschinellem Lernen, ESANN 2013. Brügge, Belgien, 24.–26. April 2013.\n",
+ "Downloaded from Kaggle:\n",
+ "https://www.kaggle.com/uciml/human-activity-recognition-with-smartphones"
]
},
{
"cell_type": "markdown",
"metadata": {
+ "Include": true,
+ "Pragraph": true,
"editable": true,
- "include": true,
- "paragraph": "DataUnderstanding",
"slideshow": {
"slide_type": ""
},
"tags": []
},
"source": [
- "Es gibt zwei getrennte Datensätze aus der Testgruppe und der Trainingsgruppe. Diese werden für die weitere Analyse zunächst noch einmal zusammengefasst. Die Daten werden identifiziert, ob sie aus dem Trainings- oder Testdatensatz stammen, damit sie später wieder aufgeteilt werden können."
+ "Es gibt zwei separate Datensätze aus der Testgruppe und der Trainingsgruppe. Zur weiteren Analyse werden diese zunächst noch einmal zusammengefasst. Die Daten werden unabhängig davon, ob sie aus dem Zug- oder Testdatensatz stammen, identifiziert, sodass sie später wieder aufgeteilt werden können."
]
},
{
"cell_type": "code",
"execution_count": 35,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"df_train = pd.read_csv('https://storage.googleapis.com/ml-service-repository-datastorage/Analysis_of_the_movement_and_activity_of_free-ranging_cattle_test.csv', delimiter=',')"
@@ -178,7 +184,14 @@
{
"cell_type": "code",
"execution_count": 36,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"df_test = pd.read_csv('https://storage.googleapis.com/ml-service-repository-datastorage/Analysis_of_the_movement_and_activity_of_free-ranging_cattle_test.csv', delimiter=',')"
@@ -187,7 +200,14 @@
{
"cell_type": "code",
"execution_count": 37,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"df_train['type']='train'"
@@ -196,7 +216,14 @@
{
"cell_type": "code",
"execution_count": 38,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"df_test['type']='test'"
@@ -205,7 +232,14 @@
{
"cell_type": "code",
"execution_count": 39,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"df = pd.concat([df_train, df_test], axis=0).reset_index(drop=True)"
@@ -214,7 +248,14 @@
{
"cell_type": "code",
"execution_count": 40,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -443,15 +484,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "Include": true,
+ "Paragraph": "Datenvorbereitung",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "3. Datenexploration und -aufbereitung"
+ "# 3. Datenvorbereitung"
]
},
{
"cell_type": "code",
"execution_count": 41,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -467,15 +523,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "Include": true,
+ "Paragraph": "Suchen nach fehlenden Daten",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Auf fehlende Werte prüfen"
+ "### Suchen nach fehlenden Daten\""
]
},
{
"cell_type": "code",
"execution_count": 42,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -506,7 +577,14 @@
{
"cell_type": "code",
"execution_count": 43,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -525,22 +603,44 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": false,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "--> **Keine fehlenden Werte im Data Frame"
+ "--> **no missing values** in the Data Frame"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "Include": true,
+ "Paragraph": "Suchen nach Duplikate",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Auf Duplikate prüfen"
+ "### Check for duplicates"
]
},
{
"cell_type": "code",
"execution_count": 44,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -611,29 +711,58 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": false,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "--> **keine Duplikatzeilen** in den Daten"
+ "--> **no duplicated rows** in the data"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "Include": true,
+ "Paragraph": "Ziel Variable",
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Ziel Variable"
+ "### Target variable"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "Wir haben ein Klassifizierungsproblem und die Zielspalte ist die Spalte \"Aktivität\"."
+ "Wir haben ein Klassifizierungsproblem und die Zielspalte ist die Spalte „Aktivität"
]
},
{
"cell_type": "code",
"execution_count": 45,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -655,7 +784,14 @@
{
"cell_type": "code",
"execution_count": 46,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -686,15 +822,29 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "* Im Smart Farming Kontext dieser Aufgabe würde ein Rind normalerweise nicht auf einer Treppe laufen. Daher werden die Zeilen mit den Bezeichnungen \"WALKING_DOWNSTAIRS\" und \"WALKING_UPSTAIRS\" entfernt:"
+ "*Im Smart-Farming-Kontext dieser Aufgabe würde ein Rind normalerweise nicht über eine Treppe laufen. Daher werden die Zeilen mit den Bezeichnungen „WALKING_DOWNSTAIRS“ und „WALKING_UPSTAIRS“ entfernt:"
]
},
{
"cell_type": "code",
"execution_count": 47,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"indexNames = df[(df['Activity'] == 'WALKING_DOWNSTAIRS') | (df['Activity'] == 'WALKING_UPSTAIRS')].index\n",
@@ -705,7 +855,14 @@
{
"cell_type": "code",
"execution_count": 48,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -736,23 +893,44 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "- Für jede Aktivität etwa die gleiche Anzahl von Beobachtungen. Wir könnten vor der Modellierung eine Überstichprobe aller Minderheitenklassen nehmen, um einen perfekt ausgewogenen Datensatz zu erhalten."
+ "- For each activity approximately the same number of observations. We could over-sample all minority classes before modelling to have a perfect balanced dataset"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "Paragraph": " Wie viele Beobachtungen gibt es zu jedem Thema?",
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Wie viele Beobachtungen gibt es von jedem Testobjekt??"
+ "### How many observations from each subject exists?"
]
},
{
"cell_type": "code",
"execution_count": 49,
"metadata": {
- "scrolled": true
+ "editable": true,
+ "include": true,
+ "scrolled": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
},
"outputs": [
{
@@ -784,22 +962,44 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "Die obige Abbildung ist interessant. Normalerweise haben alle Versuchspersonen die gleiche Versuchsreihe durchgeführt. Daher würde man annehmen, dass die Anzahl der Beobachtungen von jedem Probanden nahezu gleich sein muss. Es gibt jedoch eine Spanne von 200 bis 300 Beobachtungen über alle Versuchspersonen hinweg. Ein Grund dafür könnte sein, dass der Wechsel von einer Aktivität zur nächsten in der Sequenz nicht scharf oder deutlich genug war und die Beobachter die unklaren Beobachtungen in diesen unstabilen Phasen nachträglich gelöscht haben. "
+ "Die obige Abbildung ist interessant. Normalerweise haben alle Probanden den gleichen Versuchsablauf durchgeführt. Daher würden wir erwarten, dass die Anzahl der Beobachtungen von jedem Probanden nahezu gleich sein muss. Es gibt jedoch eine Spanne von 200 bis 300 Beobachtungen über alle Themen hinweg. Ein Grund könnte sein, dass der Übergang von einer Aktivität zur nächsten in der Sequenz nicht scharf oder klar genug war und die Beobachter die unklaren Beobachtungen in diesen instabilen Phasen nachträglich gelöscht hatten. "
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Gesamtzahl der Beobachtungen",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Gesamtzahl der Beobachtungen"
+ "### Total number of observations"
]
},
{
"cell_type": "code",
"execution_count": 50,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -815,16 +1015,31 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "PCA fuer Visionalisation",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### PCA zur Visualisierung\n",
- "- **Die PCA ist eine einfache Methode zur Visualisierung hochdimensionaler Daten in einem niedrigdimensionalen Raum. (Vorsicht: wir bezahlen dafür mit einem Informationsverlust-->aber für Visualisierungszwecke ist es OK)"
+ "### PCA fuer Visualisation\n",
+ "**Die Hauptkomponentenanalyse ** ist eine einfache Möglichkeit, hochdimensionale Daten in einem niedrigdimensionalen Raum zu visualisieren. (Achtung: Wir zahlen dafür mit einem Informationsverlust -> aber für Visualisierungszwecke ist es in Ordnung)"
]
},
{
"cell_type": "code",
"execution_count": 51,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"data_visualisation = df.drop('subject', axis =1).drop('Activity', axis=1).drop('type', axis =1)"
@@ -833,7 +1048,14 @@
{
"cell_type": "code",
"execution_count": 52,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"s = StandardScaler()\n",
@@ -843,7 +1065,14 @@
{
"cell_type": "code",
"execution_count": 53,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"#We want 3 \n",
@@ -855,7 +1084,14 @@
{
"cell_type": "code",
"execution_count": 54,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -872,7 +1108,14 @@
{
"cell_type": "code",
"execution_count": 55,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -889,7 +1132,14 @@
{
"cell_type": "code",
"execution_count": 56,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -909,7 +1159,14 @@
{
"cell_type": "code",
"execution_count": 57,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -949,7 +1206,14 @@
{
"cell_type": "code",
"execution_count": 58,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -980,7 +1244,14 @@
{
"cell_type": "code",
"execution_count": 59,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -1016,7 +1287,14 @@
{
"cell_type": "code",
"execution_count": 60,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -23187,6 +23465,26 @@
"y": 0.8999775657179135,
"z": 0.6377980341399246
},
+ "camera": {
+ "center": {
+ "x": 0,
+ "y": 0,
+ "z": 0
+ },
+ "eye": {
+ "x": 2.3987812973900966,
+ "y": 2.398781297390096,
+ "z": 2.3987812973900966
+ },
+ "projection": {
+ "type": "perspective"
+ },
+ "up": {
+ "x": 0,
+ "y": 0,
+ "z": 1
+ }
+ },
"domain": {
"x": [
0,
@@ -24075,34 +24373,61 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Ergebnisse Visualisierung mit PCA\n",
- "\n",
- "* Nach der PCA/Transformation in 3 Hauptkomponenten können wir die 3 Klassen in angemessener Weise visuell trennen. \n",
- "* Allerdings beschreiben die 3 Hauptkomponenten nur 62% der Varianz der ursprünglichen Daten. Das bedeutet, dass wir einen relativ großen Informationsverlust haben. \n",
- "* Ein zweites Problem ist, dass es schwierig ist, Modelle zu interpretieren, die auf dem Ergebnis einer PCA basieren\n",
- "\n",
+ "###Ergibnis von Visualisation mit PCA\n",
+ "Nach der PCA / Transformation in 3 Hauptkomponenten können wir die 3 Klassen visuell angemessen trennen.\n",
+ "Die 3 Hauptkomponenten beschreiben jedoch nur 62 % der Varianz der Originaldaten. Das bedeutet, dass wir einen relativ großen Informationsverlust haben.\n",
+ "Ein zweites Problem besteht darin, dass es schwierig ist, Modelle basierend auf dem Ergebnis einer PCA zu interpretieren\n",
"-->wir verwenden die PCA nur zur Visualisierung und nicht zur Modellierung"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Funktionsübersicht",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Feature Ãœbersicht"
+ "### Funktionsübersicht"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "Aus dem Originalpapier über die Daten/Quelle der Daten können wir die Information erhalten, dass es die folgenden 17 Hauptmerkmale im Zeit- und Frequenzbereich des Signals gibt:"
+ "Aus dem Originalpapier über die Daten/Quelle der Daten können wir die Information entnehmen, dass es die folgenden 17 Hauptmerkmale von Zeit- und Frequenzbereichssignalen gibt "
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
"|Name|Time|Freq.|\n",
"| --- | --- | --- |\n",
@@ -24121,7 +24446,14 @@
{
"cell_type": "code",
"execution_count": 61,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"data_temp = df.drop('subject', axis =1).drop('Activity', axis=1).drop('type', axis =1)"
@@ -24130,7 +24462,14 @@
{
"cell_type": "code",
"execution_count": 62,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -24147,7 +24486,14 @@
{
"cell_type": "code",
"execution_count": 63,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -24167,15 +24513,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "auf Multikollinearität prüfen",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Multikollinearität prüfen"
+ "### auf Multikollinearität prüfen"
]
},
{
"cell_type": "code",
"execution_count": 65,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"variables = data_temp\n",
@@ -24185,7 +24546,14 @@
{
"cell_type": "code",
"execution_count": 66,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -24786,7 +25154,14 @@
{
"cell_type": "code",
"execution_count": 67,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"# add name of features to vif dataframe\n",
@@ -24799,7 +25174,14 @@
{
"cell_type": "code",
"execution_count": 68,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -24916,7 +25298,14 @@
{
"cell_type": "code",
"execution_count": 69,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"#Drop all features with a VIF > 10, as they are correlating to much with other features\n",
@@ -24926,7 +25315,14 @@
{
"cell_type": "code",
"execution_count": 70,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"data_with_important_features = data_temp.drop(features_to_drop, axis = 1)"
@@ -24935,7 +25331,14 @@
{
"cell_type": "code",
"execution_count": 71,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -25179,7 +25582,14 @@
{
"cell_type": "code",
"execution_count": 72,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -25742,7 +26152,14 @@
{
"cell_type": "code",
"execution_count": 73,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -25778,7 +26195,14 @@
{
"cell_type": "code",
"execution_count": 74,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -25816,44 +26240,58 @@
"metadata": {
"editable": true,
"include": true,
- "paragraph": "DataUnderstanding",
+ "paragraph": "Zusammenfassung des Datenverständnisses",
"slideshow": {
"slide_type": ""
},
"tags": []
},
"source": [
- "### Zusammenfassung des Datenverständnisses\n",
- "* 1 Spalte mit der Subjekt-ID 'subject'-->nur zur Filterung, muss entfernt werden\n",
- "* 1 Zielspalte 'Aktivität'-->4 Klassen (GEHEN, SITZEN, LIEGEN, STEHEN), etwas unausgewogen\n",
- "* wir haben ein Klassifizierungsproblem mit mehreren Klassen ***überwachtes maschinelles Lernen***\n",
- "* 1 \"support column\" source-->is the observation from the train or test group-->train test split along this value\n",
- "* 1 \"Unterstützungsspalte\"-Quelle-->ist die Beobachtung aus der Trainings- oder Testgruppe-->Train-Test-Aufteilung entlang dieses Wertes \n",
- "* Diese wurden nach Möglichkeit für jede der drei Achsen x, y und z berechnet\n",
- "* In der Summe sind das 561 Merkmale\n",
- "-->Die Analyse der einzelnen Merkmale im Detail ist aufgrund der großen Anzahl von Merkmalen sehr schwierig. Schließlich haben wir Multilinerarität in den Daten\n",
+ "### Summary of the Data Understanding\n",
+ "Zusammenfassung des Datenverständnisses\n",
+ "1 Spalte mit der Betreff-ID „subject“ – nur zum Filtern – muss entfernt werden\n",
"\n",
- "* Nach der Multicollineraty-Prüfung sind nur noch 39 Merkmale übrig. In den Boxplots sind bei einigen Merkmalen Ausreißer zu sehen, aber es ist schwierig, diese Ausreißer zu analysieren und zu interpretieren, da die Daten anscheinend bereits teilweise standardisiert sind und wir keine Messeinheiten haben. Daher akzeptieren wir im Moment diese Ausreißer\n"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {},
- "source": [
- "Trainings Test Split"
+ "1 Zielspalte „Aktivität“ -> 4 Klassen (GEHEN, SITZEN, LIEGEN, STEHEN) etwas unausgewogen\n",
+ "\n",
+ "Wir haben ein Klassifizierungsproblem mit mehreren Klassen – überwachtes maschinelles Lernen\n",
+ "\n",
+ "1 Quelle „Unterstützungsspalte“ -> ist die Beobachtung aus dem Zug oder der Testgruppe -> Zugtest, aufgeteilt entlang dieses Werts\n",
+ "\n",
+ "Darüber hinaus gibt es viele Merkmale, die aus den 17 Signalmerkmalen im Zeit- und Frequenzbereich berechnet werden. Zum Beispiel Mittelwert, Min., Max., Standardwert, Signalstärkebereich usw.\n",
+ "\n",
+ "Diese wurden nach Möglichkeit für jede der drei Achsen x,y und z berechnet\n",
+ "\n",
+ "Zusammenfassend bedeutet das 561 Features -> Aufgrund der hohen Anzahl an Features ist es sehr schwierig, jedes Feature im Detail zu analysieren. Schließlich haben wir Multilinearität in den Daten\n",
+ "\n",
+ "Nach der Multikollinearitätsprüfung waren nur noch 39 Features übrig. Wir können in den Boxplots einige Ausreißer in einigen Merkmalen erkennen, aber es ist schwierig, diese Ausreißer zu analysieren und zu interpretieren, da es den Anschein hat, dass die Daten bereits teilweise standardisiert sind und wir keine Maßeinheiten haben. Im Moment akzeptieren wir diese Ausreißer\n"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph ": "Train Test Split",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### "
+ "### Train Test Split"
]
},
{
"cell_type": "code",
"execution_count": 75,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"data = df.copy()"
@@ -25862,7 +26300,14 @@
{
"cell_type": "code",
"execution_count": 76,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"data_train = data[data['type']=='train']\n",
@@ -25872,7 +26317,14 @@
{
"cell_type": "code",
"execution_count": 78,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"X_train = data_train.drop('subject', axis =1).drop('Activity', axis=1).drop('type', axis =1)\n",
@@ -25887,7 +26339,14 @@
{
"cell_type": "code",
"execution_count": 79,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -25909,15 +26368,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Data Sampling",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "### Oversampling von Daten"
+ "### Data Oversampling"
]
},
{
"cell_type": "code",
"execution_count": 80,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"ros = RandomOverSampler(random_state=0)\n",
@@ -25927,7 +26401,14 @@
{
"cell_type": "code",
"execution_count": 81,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -25958,9 +26439,17 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Modellierrung",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "# Modellierung"
+ "# Modeling "
]
},
{
@@ -25968,26 +26457,40 @@
"metadata": {
"editable": true,
"include": true,
- "paragraph": "Modeling",
"slideshow": {
"slide_type": ""
},
"tags": []
},
"source": [
- "Um verschiedene Modelle zu erstellen, ist es eine gute Praxis, Pipelines und GridSearchCV zu verwenden. Diese beiden Tools sind ein einfacher Weg, um Hyperparamter-Tuning und k-fold Cross-Validation auf verschiedenen Modellen durchzuführen. In diesem Fall bestehen die Pipelines aus einem StandardScaler und einem One vs Rest-Klassifikator"
+ "Um verschiedene Modelle zu erstellen, empfiehlt es sich, Pipelines und GridSearchCV zu verwenden. Diese beiden Tools sind eine einfache Möglichkeit, Hyperparameter-Tuning und k-fache Kreuzvalidierung auf verschiedenen Modellen durchzuführen. In diesem Fall bestehen die Pipes aus einem StandardScaler und einem One vs Rest-Klassifikator"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "__Erstellen & Evaluieren verschiedener Modelle__"
+ "__Build & Evaluate some different Models__"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph ": " KNN",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
"## KNN\n",
"K-Nearest Neighbors"
@@ -25996,7 +26499,14 @@
{
"cell_type": "code",
"execution_count": 82,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"# Pipeline for KNN\n",
@@ -26009,7 +26519,14 @@
{
"cell_type": "code",
"execution_count": 83,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -26029,7 +26546,14 @@
{
"cell_type": "code",
"execution_count": 84,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -26241,7 +26765,14 @@
{
"cell_type": "code",
"execution_count": 85,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -26260,7 +26791,14 @@
{
"cell_type": "code",
"execution_count": 86,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -26411,7 +26949,14 @@
{
"cell_type": "code",
"execution_count": 87,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"classification_report_test = classification_report(y_test, clf.best_estimator_.predict(X_test), output_dict=True)\n",
@@ -26421,7 +26966,14 @@
{
"cell_type": "code",
"execution_count": 88,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"#helper function\n",
@@ -26442,7 +26994,14 @@
{
"cell_type": "code",
"execution_count": 89,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -26615,7 +27174,14 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
"## Random Forest"
]
@@ -26623,7 +27189,14 @@
{
"cell_type": "code",
"execution_count": 90,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"# Pipeline\n",
@@ -26636,7 +27209,14 @@
{
"cell_type": "code",
"execution_count": 91,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"param_grid = {'ranfo__estimator__max_depth': [5, 15, 30]}"
@@ -26645,7 +27225,14 @@
{
"cell_type": "code",
"execution_count": 92,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -26845,7 +27432,14 @@
{
"cell_type": "code",
"execution_count": 93,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -26864,7 +27458,14 @@
{
"cell_type": "code",
"execution_count": 94,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"classification_report_test = classification_report(y_test, clf2.best_estimator_.predict(X_test), output_dict=True)\n",
@@ -26874,7 +27475,14 @@
{
"cell_type": "code",
"execution_count": 95,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27047,15 +27655,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Logistic Regression",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Logistische Regression"
+ "## Logistic Regression"
]
},
{
"cell_type": "code",
"execution_count": 96,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"# Pipeline\n",
@@ -27068,7 +27691,14 @@
{
"cell_type": "code",
"execution_count": 97,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27093,7 +27723,14 @@
{
"cell_type": "code",
"execution_count": 98,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -27112,7 +27749,14 @@
{
"cell_type": "code",
"execution_count": 99,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"classification_report_test = classification_report(y_test, clf3.best_estimator_.predict(X_test), output_dict=True)\n",
@@ -27122,7 +27766,14 @@
{
"cell_type": "code",
"execution_count": 100,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27295,15 +27946,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Decision Tree",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Entscheidungsbaum"
+ "## Decision Tree"
]
},
{
"cell_type": "code",
"execution_count": 101,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"# Pipeline\n",
@@ -27316,7 +27982,14 @@
{
"cell_type": "code",
"execution_count": 102,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27341,7 +28014,14 @@
{
"cell_type": "code",
"execution_count": 103,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -27360,7 +28040,14 @@
{
"cell_type": "code",
"execution_count": 104,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"classification_report_test = classification_report(y_test, clf4.best_estimator_.predict(X_test), output_dict=True)\n",
@@ -27370,7 +28057,15 @@
{
"cell_type": "code",
"execution_count": 105,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "SVM",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27543,7 +28238,14 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
"## SVM\n",
"Support Vector Machine"
@@ -27552,7 +28254,14 @@
{
"cell_type": "code",
"execution_count": 106,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"# Pipeline\n",
@@ -27565,7 +28274,14 @@
{
"cell_type": "code",
"execution_count": 107,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"param_grid = {\n",
@@ -27575,7 +28291,14 @@
{
"cell_type": "code",
"execution_count": 108,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27600,7 +28323,14 @@
{
"cell_type": "code",
"execution_count": 109,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"name": "stdout",
@@ -27619,7 +28349,14 @@
{
"cell_type": "code",
"execution_count": 110,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27707,7 +28444,14 @@
{
"cell_type": "code",
"execution_count": 111,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"classification_report_test = classification_report(y_test, clf5.best_estimator_.predict(X_test), output_dict=True)\n",
@@ -27717,7 +28461,14 @@
{
"cell_type": "code",
"execution_count": 112,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -27890,15 +28641,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Zusammenfassung",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Zusammenfassung"
+ "## Summary"
]
},
{
"cell_type": "code",
"execution_count": 113,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"frames = [results_KNN, \n",
@@ -27913,7 +28679,14 @@
{
"cell_type": "code",
"execution_count": 114,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -28523,7 +29296,14 @@
{
"cell_type": "code",
"execution_count": 115,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"final_results.sort_index(level=1).to_csv('Results_VIF_lower10.csv')"
@@ -28531,7 +29311,15 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Evaluation",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
"# 5. Evaluation"
]
@@ -28539,7 +29327,14 @@
{
"cell_type": "code",
"execution_count": 116,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"#helper function\n",
@@ -28571,9 +29366,17 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Genauigkeit erreicht",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Erzielte Genauigkeit"
+ "## Achieved accuracy"
]
},
{
@@ -28581,6 +29384,7 @@
"execution_count": 117,
"metadata": {
"editable": true,
+ "include": true,
"slideshow": {
"slide_type": ""
},
@@ -28689,7 +29493,14 @@
{
"cell_type": "code",
"execution_count": 118,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -28726,7 +29537,14 @@
{
"cell_type": "code",
"execution_count": 119,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -28882,32 +29700,46 @@
"metadata": {
"editable": true,
"include": true,
- "paragraph": "Evaluation",
"slideshow": {
"slide_type": ""
},
"tags": []
},
"source": [
- "* **KNN** --> Die Leistung auf Testdaten ist ca. 8% niedriger als auf Trainingsdaten. Wahrscheinlich passt sich das Modell zu sehr an die Trainingsdaten an und lässt sich nicht gut auf ungesehene Testdaten verallgemeinern.\n",
- "* **Random Forest** --> Die Leistung bei den Testdaten ist 15% geringer als bei den Trainingsdaten. Das Modell **übererfüllt** die Trainingsdaten und verallgemeinert sich nicht gut auf die ungesehenen Testdaten.\n",
- "* **Logistische Regression** --> Trainings- und Testleistung sind sehr ähnlich (diff=3.9%). Dies bedeutet wahrscheinlich, dass das erstellte Modell **gut** auf neue Daten verallgemeinert.\n",
- "* **Entscheidungsbaum** --> Die Leistung bei den Testdaten ist 34% niedriger als bei den Trainingsdaten. Das Modell passt sich zu sehr an die Trainingsdaten an und verallgemeinert **nicht gut** auf die unbekannten Testdaten.\n",
- "* **Support Vector Machine** --> Trainings- und Testleistung sind sehr ähnlich (diff=4.8%). Das bedeutet wahrscheinlich, dass das erstellte Modell **gut** auf neue Daten verallgemeinert.\n",
- "-->Überprüfen wir die Konfusionsmatrix der beiden besten Modelle"
+ "KNN -> Die Leistung bei Testdaten beträgt ca. 8 % niedriger als bei Trainingsdaten. Wahrscheinlich passt das Modell zu stark auf Trainingsdaten und lässt sich nicht gut auf nicht sichtbare Testdaten verallgemeinern.\n",
+ "Random Forest -> Die Leistung bei Testdaten ist 15 % geringer als bei Trainingsdaten. Das Modell passt zu stark auf Trainingsdaten und lässt sich nicht gut auf nicht sichtbare Testdaten verallgemeinern.\n",
+ "Logistische Regression -> Trainings- und Testleistung sind sehr ähnlich (Differenz = 3,9 %). Dies bedeutet wahrscheinlich, dass sich das erstellte Modell gut auf neue Daten verallgemeinern lässt.\n",
+ "Entscheidungsbaum -> Die Leistung bei Testdaten ist 34 % geringer als bei Trainingsdaten. Das Modell passt zu stark auf Trainingsdaten und lässt sich nicht gut auf nicht sichtbare Testdaten verallgemeinern.\n",
+ "Support Vector Machine -> Trainings- und Testleistung sind sehr ähnlich (Differenz = 4,8 %). Dies bedeutet wahrscheinlich, dass sich das erstellte Modell gut auf neue Daten verallgemeinern lässt.\n",
+ "-->Lassen Sie uns die Verwirrungsmatrix der beiden besten Modelle überprüfen"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragragh": "Confusion Matrix-Logistic Regression",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Konfusionsmatrix-Logistische Regression"
+ "## Confusion Matrix-Logistic Regression"
]
},
{
"cell_type": "code",
"execution_count": 120,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -28928,15 +29760,30 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Confusion Matrix SVM",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Konfusionsmatrix-SVM"
+ "## Confusion Matrix-SVM"
]
},
{
"cell_type": "code",
"execution_count": 121,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -28959,7 +29806,13 @@
"cell_type": "code",
"execution_count": 122,
"metadata": {
- "scrolled": true
+ "editable": true,
+ "include": true,
+ "scrolled": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
},
"outputs": [
{
@@ -29143,18 +29996,32 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragrph": "Funktionsbedeutung",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Wichtigkeit der Merkmale\n",
+ "## Funktionsbedeutung\n",
"\n",
- "* \n",
- "Ein One vs Rest Classifier besteht aus n einzelnen Klassifikatoren (mit n=Anzahl der Klassen). Das macht es schwierig, die Merkmalsbedeutung für den gesamten OvsR-Klassifikator zu interpretieren. Eine Möglichkeit besteht darin, die Koeffizienten für jedes einzelne Modell zu bestimmen und den Mittelwert über jeden Koeffizienten zu bilden.\n"
+ "* Ein One vs Rest-Klassifikator besteht aus n einzelnen Klassifikatoren (wobei n = Anzahl der Klassen). Das macht es schwierig, die Bedeutung der Merkmale für den gesamten OvsR-Klassifikator zu interpretieren. Eine Möglichkeit besteht darin, die Koeffizienten für jedes einzelne Modell zu bestimmen und den Mittelwert über jeden Koeffizienten zu bilden."
]
},
{
"cell_type": "code",
"execution_count": 123,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"def get_the_feature_importances(clfobject,est_name, functype):\n",
@@ -29174,7 +30041,14 @@
{
"cell_type": "code",
"execution_count": 124,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"feature_importance = pd.DataFrame({},[])\n",
@@ -29184,18 +30058,33 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "jp-MarkdownHeadingCollapsed": true,
+ "paragraph": "ergebnis",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Ergebnisse"
+ "## Results"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "* Das beste Modell ist der OvsR-Klassifikator mit logistischer Regression. \n",
- "* Es hat eine gute Genauigkeit (0.77:test, 0.81:train) und eine gute Präzision (0.87:test) für die Klasse 'WALKING'.\n",
- "* Die 10 wichtigsten Eigenschaften dieses OvsR-Klassifikators sind die folgenden:"
+ "* The best model is the Logistic Regression OvsR Classifier. \n",
+ "* It has a good accuracy (0.77:test, 0.81:train) and a good precision (0.87:test) for the class 'WALKING'.\n",
+ "* The 10 most important feautures of this OvsR-Classifier are the following:"
]
},
{
@@ -29312,20 +30201,35 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "-->Die meisten dieser wichtigen Funktionen stammen vom Gyroskopsensor im Smartphone."
+ "-->The most of this important features comes from the **gyroscope sensor** in the smartphone"
]
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Modell Deployment",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
- "## Endgültiges Modell für das Deployment"
+ "## Final model for deployment"
]
},
{
"cell_type": "code",
"execution_count": 126,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [],
"source": [
"final_model = clf3.best_estimator_"
@@ -29333,16 +30237,31 @@
},
{
"cell_type": "markdown",
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "paragraph": "Deployment",
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"source": [
"# 6. Deployment\n",
- "Teste die Prognose lokal"
+ "Test the prediction locally:"
]
},
{
"cell_type": "code",
"execution_count": 127,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -29561,7 +30480,14 @@
{
"cell_type": "code",
"execution_count": 128,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -29586,7 +30512,14 @@
{
"cell_type": "code",
"execution_count": 129,
- "metadata": {},
+ "metadata": {
+ "editable": true,
+ "include": true,
+ "slideshow": {
+ "slide_type": ""
+ },
+ "tags": []
+ },
"outputs": [
{
"data": {
@@ -29608,7 +30541,6 @@
"metadata": {
"editable": true,
"include": true,
- "paragraph": "Deployment",
"slideshow": {
"slide_type": ""
},
@@ -29621,7 +30553,8 @@
],
"metadata": {
"branche": "Landwirtschaft",
- "dataSource": "https://storage.googleapis.com/ml-service-repository-datastorage/Analysis_of_the_movement_and_activity_of_free-ranging_cattle_test.csv",
+ "category": "Forschung",
+ "dataSource": "https://archive.ics.uci.edu/ml/datasets/human+activity+recognition+using+smartphones",
"funktion": "Operations",
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
@@ -29640,10 +30573,10 @@
"pygments_lexer": "ipython3",
"version": "3.12.3"
},
- "repoLink": "https://gitlab.reutlingen-university.de/ki_lab/machine-learning-services/-/tree/main/Agriculture/Analysis%20of%20the%20movement%20and%20activity%20of%20free-ranging%20cattle",
+ "repoLink": "https://github.com/AlexRossmann/machine-learning-services/tree/main/Agriculture/Analysis%20of%20the%20movement%20and%20activity%20of%20free-ranging%20cattle",
"skipNotebookInDeployment": false,
- "teaser": "To be done",
- "title": "Analyse der Bewegungsprofile von Kühen auf dem Feld"
+ "teaser": "Landwirte geraten immer weiter unter wirtschaftlichen Druck. Beispielsweise kam es durch die Coronapandemie zu teuren logistischen Problemen. Gleichzeitig stellen immer mehr Konsumenten höhere Anforderungen an eine artgerechte Tierhaltung. Diese Marktbedingungen erscheinen für Landwirte problematisch, da letzterer Aspekt natürlich mit gewissen Kosten verbunden ist und ein generelles Umdenken notwendig macht ",
+ "title": "Erkennung der Aktivitäten von Kühen auf dem Feld"
},
"nbformat": 4,
"nbformat_minor": 4
diff --git a/Tourism/Prediction cancellation of hotel bookings/notebook.ipynb b/Tourism/Prediction cancellation of hotel bookings/notebook.ipynb
deleted file mode 100644
index b03d30e..0000000
--- a/Tourism/Prediction cancellation of hotel bookings/notebook.ipynb
+++ /dev/null
@@ -1,2641 +0,0 @@
-{
- "cells": [
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Geschäftsverständnis",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "# 1. Geschäftsverständnis"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "Abschätzung des Kundenverhaltens bezüglich Hotelstornierungen zur Planung von Kapazitäten. Der Anwendungsfall soll testen, ob es möglich ist,\n",
- "Hotelstornierungen vorherzusagen.Ein wichtiges Ziel für jedes Unternehmen liegt in der Erhaltung wertvoller Kundenbeziehungen, \n",
- "postive Ranking und Kapazitat plannung. Für Unternehmen ist daher eine Einschätzung der Kundenvehalten bei Buchungstonierung notwendig, \n",
- "So dass sich das Risiko der Buchungstonierung eines Kunden vorab einschätzen lässt, können Gegenmaßnahmen eingeleitet werden."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Daten und Datenverständnis",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "# 2. Daten und Datenverständnis"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "Die Daten Rahmen sind fuer Hotel Buchungen und Der Datensatz für diese Demo wurde auf der Kaggle Data Science Plattform als c.s.v file veröffentlicht. \n",
- "In den Datensätzen sind allerlei Daten zu den Gästen erfasst. Merkmale zu Familie mit Kindern Buchungen über Reisebüros etc. können Aufschluss \n",
- "darüber geben, ob bei ihnen eine höhere Stornoquote vorliegt. Der Datenrahmen enthält Buchungsinformationen von 2 verschiedenen Hotels und die\n",
- "Anzahl der Beobachtungen = 119390, Anzahl der Merkmale = 32\n",
- "Correlation Analysis: stays in weekend nights and stays in week night with 0.5."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Import von Relevant Module",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "# 2.1. Import von Relevant Module"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 2,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "import pandas as pd\n",
- "import numpy as np\n",
- "import matplotlib.pyplot as plt\n",
- "import seaborn as sns\n",
- "sns.set()\n",
- "\n",
- "from sklearn.model_selection import train_test_split\n",
- "from sklearn.tree import DecisionTreeClassifier\n",
- "from sklearn.ensemble import RandomForestClassifier\n",
- "from sklearn.linear_model import LogisticRegression\n",
- "from sklearn import metrics\n",
- "from sklearn.metrics import confusion_matrix, classification_report"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Daten Auslesen",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 2.2. Daten Auslesen"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 3,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>hotel</th>\n",
- " <th>is_canceled</th>\n",
- " <th>lead_time</th>\n",
- " <th>arrival_date_year</th>\n",
- " <th>arrival_date_month</th>\n",
- " <th>arrival_date_week_number</th>\n",
- " <th>arrival_date_day_of_month</th>\n",
- " <th>stays_in_weekend_nights</th>\n",
- " <th>stays_in_week_nights</th>\n",
- " <th>adults</th>\n",
- " <th>...</th>\n",
- " <th>deposit_type</th>\n",
- " <th>agent</th>\n",
- " <th>company</th>\n",
- " <th>days_in_waiting_list</th>\n",
- " <th>customer_type</th>\n",
- " <th>adr</th>\n",
- " <th>required_car_parking_spaces</th>\n",
- " <th>total_of_special_requests</th>\n",
- " <th>reservation_status</th>\n",
- " <th>reservation_status_date</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <td>0</td>\n",
- " <td>Resort Hotel</td>\n",
- " <td>0</td>\n",
- " <td>342</td>\n",
- " <td>2015</td>\n",
- " <td>July</td>\n",
- " <td>27</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>...</td>\n",
- " <td>No Deposit</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>0</td>\n",
- " <td>Transient</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>Check-Out</td>\n",
- " <td>2015-07-01</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>1</td>\n",
- " <td>Resort Hotel</td>\n",
- " <td>0</td>\n",
- " <td>737</td>\n",
- " <td>2015</td>\n",
- " <td>July</td>\n",
- " <td>27</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>...</td>\n",
- " <td>No Deposit</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>0</td>\n",
- " <td>Transient</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>Check-Out</td>\n",
- " <td>2015-07-01</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>2</td>\n",
- " <td>Resort Hotel</td>\n",
- " <td>0</td>\n",
- " <td>7</td>\n",
- " <td>2015</td>\n",
- " <td>July</td>\n",
- " <td>27</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>No Deposit</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>0</td>\n",
- " <td>Transient</td>\n",
- " <td>75.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>Check-Out</td>\n",
- " <td>2015-07-02</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>3</td>\n",
- " <td>Resort Hotel</td>\n",
- " <td>0</td>\n",
- " <td>13</td>\n",
- " <td>2015</td>\n",
- " <td>July</td>\n",
- " <td>27</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>No Deposit</td>\n",
- " <td>304.0</td>\n",
- " <td>NaN</td>\n",
- " <td>0</td>\n",
- " <td>Transient</td>\n",
- " <td>75.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>Check-Out</td>\n",
- " <td>2015-07-02</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>4</td>\n",
- " <td>Resort Hotel</td>\n",
- " <td>0</td>\n",
- " <td>14</td>\n",
- " <td>2015</td>\n",
- " <td>July</td>\n",
- " <td>27</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>2</td>\n",
- " <td>...</td>\n",
- " <td>No Deposit</td>\n",
- " <td>240.0</td>\n",
- " <td>NaN</td>\n",
- " <td>0</td>\n",
- " <td>Transient</td>\n",
- " <td>98.0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>Check-Out</td>\n",
- " <td>2015-07-03</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "<p>5 rows × 32 columns</p>\n",
- "</div>"
- ],
- "text/plain": [
- " hotel is_canceled lead_time arrival_date_year arrival_date_month \\\n",
- "0 Resort Hotel 0 342 2015 July \n",
- "1 Resort Hotel 0 737 2015 July \n",
- "2 Resort Hotel 0 7 2015 July \n",
- "3 Resort Hotel 0 13 2015 July \n",
- "4 Resort Hotel 0 14 2015 July \n",
- "\n",
- " arrival_date_week_number arrival_date_day_of_month \\\n",
- "0 27 1 \n",
- "1 27 1 \n",
- "2 27 1 \n",
- "3 27 1 \n",
- "4 27 1 \n",
- "\n",
- " stays_in_weekend_nights stays_in_week_nights adults ... deposit_type \\\n",
- "0 0 0 2 ... No Deposit \n",
- "1 0 0 2 ... No Deposit \n",
- "2 0 1 1 ... No Deposit \n",
- "3 0 1 1 ... No Deposit \n",
- "4 0 2 2 ... No Deposit \n",
- "\n",
- " agent company days_in_waiting_list customer_type adr \\\n",
- "0 NaN NaN 0 Transient 0.0 \n",
- "1 NaN NaN 0 Transient 0.0 \n",
- "2 NaN NaN 0 Transient 75.0 \n",
- "3 304.0 NaN 0 Transient 75.0 \n",
- "4 240.0 NaN 0 Transient 98.0 \n",
- "\n",
- " required_car_parking_spaces total_of_special_requests reservation_status \\\n",
- "0 0 0 Check-Out \n",
- "1 0 0 Check-Out \n",
- "2 0 0 Check-Out \n",
- "3 0 0 Check-Out \n",
- "4 0 1 Check-Out \n",
- "\n",
- " reservation_status_date \n",
- "0 2015-07-01 \n",
- "1 2015-07-01 \n",
- "2 2015-07-02 \n",
- "3 2015-07-02 \n",
- "4 2015-07-03 \n",
- "\n",
- "[5 rows x 32 columns]"
- ]
- },
- "execution_count": 3,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "# my_file = project.get_file()\n",
- "\n",
- "# my_file.seek(0)\n",
- "df = pd.read_csv(\"https://storage.googleapis.com/ml-service-repository-datastorage/Prediction_cancellation_of_hotel_bookings_data.csv\")\n",
- "\n",
- "df.head()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 4,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "def attribute_description(data):\n",
- " longestColumnName = len(max(np.array(data.columns), key=len))\n",
- " for col in data.columns:\n",
- " description = ''\n",
- " col_dropna = data[col].dropna()\n",
- " example = col_dropna.sample(1).values[0]\n",
- " if type(example) == str:\n",
- " description = 'str '\n",
- " if len(col_dropna.unique()) < 10:\n",
- " description += '['\n",
- " description += '; '.join([ f'\"{name}\"' for name in col_dropna.unique()])\n",
- " description += ']'\n",
- " else:\n",
- " description += '[ example: \"'+ example + '\" ]'\n",
- " else:\n",
- " description = str(type(example))\n",
- " print(col.ljust(longestColumnName)+ f': {description}')"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 5,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "hotel : str [\"Resort Hotel\"; \"City Hotel\"]\n",
- "is_canceled : <class 'numpy.int64'>\n",
- "lead_time : <class 'numpy.int64'>\n",
- "arrival_date_year : <class 'numpy.int64'>\n",
- "arrival_date_month : str [ example: \"May\" ]\n",
- "arrival_date_week_number : <class 'numpy.int64'>\n",
- "arrival_date_day_of_month : <class 'numpy.int64'>\n",
- "stays_in_weekend_nights : <class 'numpy.int64'>\n",
- "stays_in_week_nights : <class 'numpy.int64'>\n",
- "adults : <class 'numpy.int64'>\n",
- "children : <class 'numpy.float64'>\n",
- "babies : <class 'numpy.int64'>\n",
- "meal : str [\"BB\"; \"FB\"; \"HB\"; \"SC\"; \"Undefined\"]\n",
- "country : str [ example: \"SWE\" ]\n",
- "market_segment : str [\"Direct\"; \"Corporate\"; \"Online TA\"; \"Offline TA/TO\"; \"Complementary\"; \"Groups\"; \"Undefined\"; \"Aviation\"]\n",
- "distribution_channel : str [\"Direct\"; \"Corporate\"; \"TA/TO\"; \"Undefined\"; \"GDS\"]\n",
- "is_repeated_guest : <class 'numpy.int64'>\n",
- "previous_cancellations : <class 'numpy.int64'>\n",
- "previous_bookings_not_canceled: <class 'numpy.int64'>\n",
- "reserved_room_type : str [ example: \"A\" ]\n",
- "assigned_room_type : str [ example: \"C\" ]\n",
- "booking_changes : <class 'numpy.int64'>\n",
- "deposit_type : str [\"No Deposit\"; \"Refundable\"; \"Non Refund\"]\n",
- "agent : <class 'numpy.float64'>\n",
- "company : <class 'numpy.float64'>\n",
- "days_in_waiting_list : <class 'numpy.int64'>\n",
- "customer_type : str [\"Transient\"; \"Contract\"; \"Transient-Party\"; \"Group\"]\n",
- "adr : <class 'numpy.float64'>\n",
- "required_car_parking_spaces : <class 'numpy.int64'>\n",
- "total_of_special_requests : <class 'numpy.int64'>\n",
- "reservation_status : str [\"Check-Out\"; \"Canceled\"; \"No-Show\"]\n",
- "reservation_status_date : str [ example: \"2017-02-27\" ]\n"
- ]
- }
- ],
- "source": [
- "attribute_description(df)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 6,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>hotel</th>\n",
- " <th>is_canceled</th>\n",
- " <th>lead_time</th>\n",
- " <th>arrival_date_year</th>\n",
- " <th>arrival_date_month</th>\n",
- " <th>arrival_date_week_number</th>\n",
- " <th>arrival_date_day_of_month</th>\n",
- " <th>stays_in_weekend_nights</th>\n",
- " <th>stays_in_week_nights</th>\n",
- " <th>adults</th>\n",
- " <th>...</th>\n",
- " <th>deposit_type</th>\n",
- " <th>agent</th>\n",
- " <th>company</th>\n",
- " <th>days_in_waiting_list</th>\n",
- " <th>customer_type</th>\n",
- " <th>adr</th>\n",
- " <th>required_car_parking_spaces</th>\n",
- " <th>total_of_special_requests</th>\n",
- " <th>reservation_status</th>\n",
- " <th>reservation_status_date</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <td>count</td>\n",
- " <td>119390</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>...</td>\n",
- " <td>119390</td>\n",
- " <td>103050.000000</td>\n",
- " <td>6797.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390.000000</td>\n",
- " <td>119390</td>\n",
- " <td>119390</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>unique</td>\n",
- " <td>2</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>12</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>...</td>\n",
- " <td>3</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>4</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>3</td>\n",
- " <td>926</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>top</td>\n",
- " <td>City Hotel</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>August</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>...</td>\n",
- " <td>No Deposit</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>Transient</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>Check-Out</td>\n",
- " <td>2015-10-21</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>freq</td>\n",
- " <td>79330</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>13877</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>...</td>\n",
- " <td>104641</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>89613</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " <td>75166</td>\n",
- " <td>1461</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>mean</td>\n",
- " <td>NaN</td>\n",
- " <td>0.370416</td>\n",
- " <td>104.011416</td>\n",
- " <td>2016.156554</td>\n",
- " <td>NaN</td>\n",
- " <td>27.165173</td>\n",
- " <td>15.798241</td>\n",
- " <td>0.927599</td>\n",
- " <td>2.500302</td>\n",
- " <td>1.856403</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>86.693382</td>\n",
- " <td>189.266735</td>\n",
- " <td>2.321149</td>\n",
- " <td>NaN</td>\n",
- " <td>101.831122</td>\n",
- " <td>0.062518</td>\n",
- " <td>0.571363</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>std</td>\n",
- " <td>NaN</td>\n",
- " <td>0.482918</td>\n",
- " <td>106.863097</td>\n",
- " <td>0.707476</td>\n",
- " <td>NaN</td>\n",
- " <td>13.605138</td>\n",
- " <td>8.780829</td>\n",
- " <td>0.998613</td>\n",
- " <td>1.908286</td>\n",
- " <td>0.579261</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>110.774548</td>\n",
- " <td>131.655015</td>\n",
- " <td>17.594721</td>\n",
- " <td>NaN</td>\n",
- " <td>50.535790</td>\n",
- " <td>0.245291</td>\n",
- " <td>0.792798</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>min</td>\n",
- " <td>NaN</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>2015.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>1.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>1.000000</td>\n",
- " <td>6.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>-6.380000</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>25%</td>\n",
- " <td>NaN</td>\n",
- " <td>0.000000</td>\n",
- " <td>18.000000</td>\n",
- " <td>2016.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>16.000000</td>\n",
- " <td>8.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>2.000000</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>9.000000</td>\n",
- " <td>62.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>69.290000</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>50%</td>\n",
- " <td>NaN</td>\n",
- " <td>0.000000</td>\n",
- " <td>69.000000</td>\n",
- " <td>2016.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>28.000000</td>\n",
- " <td>16.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>2.000000</td>\n",
- " <td>2.000000</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>14.000000</td>\n",
- " <td>179.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>94.575000</td>\n",
- " <td>0.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>75%</td>\n",
- " <td>NaN</td>\n",
- " <td>1.000000</td>\n",
- " <td>160.000000</td>\n",
- " <td>2017.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>38.000000</td>\n",
- " <td>23.000000</td>\n",
- " <td>2.000000</td>\n",
- " <td>3.000000</td>\n",
- " <td>2.000000</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>229.000000</td>\n",
- " <td>270.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>126.000000</td>\n",
- " <td>0.000000</td>\n",
- " <td>1.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>max</td>\n",
- " <td>NaN</td>\n",
- " <td>1.000000</td>\n",
- " <td>737.000000</td>\n",
- " <td>2017.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>53.000000</td>\n",
- " <td>31.000000</td>\n",
- " <td>19.000000</td>\n",
- " <td>50.000000</td>\n",
- " <td>55.000000</td>\n",
- " <td>...</td>\n",
- " <td>NaN</td>\n",
- " <td>535.000000</td>\n",
- " <td>543.000000</td>\n",
- " <td>391.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>5400.000000</td>\n",
- " <td>8.000000</td>\n",
- " <td>5.000000</td>\n",
- " <td>NaN</td>\n",
- " <td>NaN</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "<p>11 rows × 32 columns</p>\n",
- "</div>"
- ],
- "text/plain": [
- " hotel is_canceled lead_time arrival_date_year \\\n",
- "count 119390 119390.000000 119390.000000 119390.000000 \n",
- "unique 2 NaN NaN NaN \n",
- "top City Hotel NaN NaN NaN \n",
- "freq 79330 NaN NaN NaN \n",
- "mean NaN 0.370416 104.011416 2016.156554 \n",
- "std NaN 0.482918 106.863097 0.707476 \n",
- "min NaN 0.000000 0.000000 2015.000000 \n",
- "25% NaN 0.000000 18.000000 2016.000000 \n",
- "50% NaN 0.000000 69.000000 2016.000000 \n",
- "75% NaN 1.000000 160.000000 2017.000000 \n",
- "max NaN 1.000000 737.000000 2017.000000 \n",
- "\n",
- " arrival_date_month arrival_date_week_number \\\n",
- "count 119390 119390.000000 \n",
- "unique 12 NaN \n",
- "top August NaN \n",
- "freq 13877 NaN \n",
- "mean NaN 27.165173 \n",
- "std NaN 13.605138 \n",
- "min NaN 1.000000 \n",
- "25% NaN 16.000000 \n",
- "50% NaN 28.000000 \n",
- "75% NaN 38.000000 \n",
- "max NaN 53.000000 \n",
- "\n",
- " arrival_date_day_of_month stays_in_weekend_nights \\\n",
- "count 119390.000000 119390.000000 \n",
- "unique NaN NaN \n",
- "top NaN NaN \n",
- "freq NaN NaN \n",
- "mean 15.798241 0.927599 \n",
- "std 8.780829 0.998613 \n",
- "min 1.000000 0.000000 \n",
- "25% 8.000000 0.000000 \n",
- "50% 16.000000 1.000000 \n",
- "75% 23.000000 2.000000 \n",
- "max 31.000000 19.000000 \n",
- "\n",
- " stays_in_week_nights adults ... deposit_type agent \\\n",
- "count 119390.000000 119390.000000 ... 119390 103050.000000 \n",
- "unique NaN NaN ... 3 NaN \n",
- "top NaN NaN ... No Deposit NaN \n",
- "freq NaN NaN ... 104641 NaN \n",
- "mean 2.500302 1.856403 ... NaN 86.693382 \n",
- "std 1.908286 0.579261 ... NaN 110.774548 \n",
- "min 0.000000 0.000000 ... NaN 1.000000 \n",
- "25% 1.000000 2.000000 ... NaN 9.000000 \n",
- "50% 2.000000 2.000000 ... NaN 14.000000 \n",
- "75% 3.000000 2.000000 ... NaN 229.000000 \n",
- "max 50.000000 55.000000 ... NaN 535.000000 \n",
- "\n",
- " company days_in_waiting_list customer_type adr \\\n",
- "count 6797.000000 119390.000000 119390 119390.000000 \n",
- "unique NaN NaN 4 NaN \n",
- "top NaN NaN Transient NaN \n",
- "freq NaN NaN 89613 NaN \n",
- "mean 189.266735 2.321149 NaN 101.831122 \n",
- "std 131.655015 17.594721 NaN 50.535790 \n",
- "min 6.000000 0.000000 NaN -6.380000 \n",
- "25% 62.000000 0.000000 NaN 69.290000 \n",
- "50% 179.000000 0.000000 NaN 94.575000 \n",
- "75% 270.000000 0.000000 NaN 126.000000 \n",
- "max 543.000000 391.000000 NaN 5400.000000 \n",
- "\n",
- " required_car_parking_spaces total_of_special_requests \\\n",
- "count 119390.000000 119390.000000 \n",
- "unique NaN NaN \n",
- "top NaN NaN \n",
- "freq NaN NaN \n",
- "mean 0.062518 0.571363 \n",
- "std 0.245291 0.792798 \n",
- "min 0.000000 0.000000 \n",
- "25% 0.000000 0.000000 \n",
- "50% 0.000000 0.000000 \n",
- "75% 0.000000 1.000000 \n",
- "max 8.000000 5.000000 \n",
- "\n",
- " reservation_status reservation_status_date \n",
- "count 119390 119390 \n",
- "unique 3 926 \n",
- "top Check-Out 2015-10-21 \n",
- "freq 75166 1461 \n",
- "mean NaN NaN \n",
- "std NaN NaN \n",
- "min NaN NaN \n",
- "25% NaN NaN \n",
- "50% NaN NaN \n",
- "75% NaN NaN \n",
- "max NaN NaN \n",
- "\n",
- "[11 rows x 32 columns]"
- ]
- },
- "execution_count": 6,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "df.describe(include='all')"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Daten Vorbereitung",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 2.3. Daten bereingung"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 7,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1296x1296 with 2 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "f,ax=plt.subplots(figsize = (18,18))\n",
- "sns.heatmap(df.corr(),annot= True,linewidths=0.5,fmt = \".1f\",ax=ax)\n",
- "plt.xticks(rotation=90)\n",
- "plt.yticks(rotation=0)\n",
- "plt.title('Correlation Map')\n",
- "plt.show()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 8,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/plain": [
- "hotel 0\n",
- "is_canceled 0\n",
- "lead_time 0\n",
- "arrival_date_year 0\n",
- "arrival_date_month 0\n",
- "arrival_date_week_number 0\n",
- "arrival_date_day_of_month 0\n",
- "stays_in_weekend_nights 0\n",
- "stays_in_week_nights 0\n",
- "adults 0\n",
- "children 4\n",
- "babies 0\n",
- "meal 0\n",
- "country 488\n",
- "market_segment 0\n",
- "distribution_channel 0\n",
- "is_repeated_guest 0\n",
- "previous_cancellations 0\n",
- "previous_bookings_not_canceled 0\n",
- "reserved_room_type 0\n",
- "assigned_room_type 0\n",
- "booking_changes 0\n",
- "deposit_type 0\n",
- "agent 16340\n",
- "company 112593\n",
- "days_in_waiting_list 0\n",
- "customer_type 0\n",
- "adr 0\n",
- "required_car_parking_spaces 0\n",
- "total_of_special_requests 0\n",
- "reservation_status 0\n",
- "reservation_status_date 0\n",
- "dtype: int64"
- ]
- },
- "execution_count": 8,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "df.isnull().sum()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 9,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['reservation_status'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 10,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['stays_in_weekend_nights'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 11,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['reservation_status_date'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 12,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['arrival_date_day_of_month'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 13,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['arrival_date_year'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 14,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['arrival_date_month'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 15,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['arrival_date_week_number'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 16,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['required_car_parking_spaces'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 17,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['previous_bookings_not_canceled'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 18,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['total_of_special_requests'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 19,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['agent'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 20,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['company'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 21,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.drop(['adr'], axis=1)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 22,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "df = df.dropna(axis=0)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Test auf Multikollinearität",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 2.4. Test for Multicollinearity"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 23,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
- "variables = df[['lead_time', 'is_repeated_guest', 'adults', 'booking_changes', 'previous_cancellations', 'is_canceled', 'stays_in_week_nights', 'babies', 'days_in_waiting_list']]\n",
- "vif = pd.DataFrame()\n",
- "vif['VIF'] = [variance_inflation_factor(variables.values, i) for i in range(variables.shape[1])]\n",
- "vif['Features'] = variables.columns"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 24,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>VIF</th>\n",
- " <th>Features</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <td>0</td>\n",
- " <td>2.285568</td>\n",
- " <td>lead_time</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>1</td>\n",
- " <td>1.033605</td>\n",
- " <td>is_repeated_guest</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>2</td>\n",
- " <td>3.354523</td>\n",
- " <td>adults</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>3</td>\n",
- " <td>1.143147</td>\n",
- " <td>booking_changes</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>4</td>\n",
- " <td>1.037159</td>\n",
- " <td>previous_cancellations</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>5</td>\n",
- " <td>1.759416</td>\n",
- " <td>is_canceled</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>6</td>\n",
- " <td>2.680081</td>\n",
- " <td>stays_in_week_nights</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>7</td>\n",
- " <td>1.015332</td>\n",
- " <td>babies</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>8</td>\n",
- " <td>1.049190</td>\n",
- " <td>days_in_waiting_list</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "</div>"
- ],
- "text/plain": [
- " VIF Features\n",
- "0 2.285568 lead_time\n",
- "1 1.033605 is_repeated_guest\n",
- "2 3.354523 adults\n",
- "3 1.143147 booking_changes\n",
- "4 1.037159 previous_cancellations\n",
- "5 1.759416 is_canceled\n",
- "6 2.680081 stays_in_week_nights\n",
- "7 1.015332 babies\n",
- "8 1.049190 days_in_waiting_list"
- ]
- },
- "execution_count": 24,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "vif"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Deskriptiv Analyse",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 2.5. Deskriptive Analyse"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 25,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/plain": [
- "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37350C08>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37C47A88>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37C572C8>],\n",
- " [<matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37C913C8>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37CC8508>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37D02608>],\n",
- " [<matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37D3B708>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37D73808>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37D80408>],\n",
- " [<matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37DB85C8>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37E1EB48>,\n",
- " <matplotlib.axes._subplots.AxesSubplot object at 0x000001DE37E54D08>]],\n",
- " dtype=object)"
- ]
- },
- "execution_count": 25,
- "metadata": {},
- "output_type": "execute_result"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 1800x1800 with 12 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "df.hist(figsize=(25,25), bins=50)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Datenaufbereitung",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "# 3. Datenaufbereitung"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "Zunächst wird der Typ der Daten nach dem Einlesen in das Notebook überprüft. Einlesefehler werden entsprechend korrigiert.\n",
- "Dimensionalitäts reduktion: entfernte Attribute ohne Beschreibung. Fehlende Daten: Zeilen mit fehlenden Daten werden entfernt.\n",
- "Datenkonvertierung: Dummy-Variablen werden erstellt."
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Erfassung kategorialer Variablen",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 3.1. Recoding of Categorical Variables"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 26,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>is_canceled</th>\n",
- " <th>lead_time</th>\n",
- " <th>stays_in_week_nights</th>\n",
- " <th>adults</th>\n",
- " <th>children</th>\n",
- " <th>babies</th>\n",
- " <th>is_repeated_guest</th>\n",
- " <th>previous_cancellations</th>\n",
- " <th>booking_changes</th>\n",
- " <th>days_in_waiting_list</th>\n",
- " <th>...</th>\n",
- " <th>assigned_room_type_H</th>\n",
- " <th>assigned_room_type_I</th>\n",
- " <th>assigned_room_type_K</th>\n",
- " <th>assigned_room_type_L</th>\n",
- " <th>assigned_room_type_P</th>\n",
- " <th>deposit_type_Non Refund</th>\n",
- " <th>deposit_type_Refundable</th>\n",
- " <th>customer_type_Group</th>\n",
- " <th>customer_type_Transient</th>\n",
- " <th>customer_type_Transient-Party</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>342</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>3</td>\n",
- " <td>0</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " <td>737</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>4</td>\n",
- " <td>0</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>2</td>\n",
- " <td>0</td>\n",
- " <td>7</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>3</td>\n",
- " <td>0</td>\n",
- " <td>13</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>4</td>\n",
- " <td>0</td>\n",
- " <td>14</td>\n",
- " <td>2</td>\n",
- " <td>2</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "<p>5 rows × 226 columns</p>\n",
- "</div>"
- ],
- "text/plain": [
- " is_canceled lead_time stays_in_week_nights adults children babies \\\n",
- "0 0 342 0 2 0.0 0 \n",
- "1 0 737 0 2 0.0 0 \n",
- "2 0 7 1 1 0.0 0 \n",
- "3 0 13 1 1 0.0 0 \n",
- "4 0 14 2 2 0.0 0 \n",
- "\n",
- " is_repeated_guest previous_cancellations booking_changes \\\n",
- "0 0 0 3 \n",
- "1 0 0 4 \n",
- "2 0 0 0 \n",
- "3 0 0 0 \n",
- "4 0 0 0 \n",
- "\n",
- " days_in_waiting_list ... assigned_room_type_H assigned_room_type_I \\\n",
- "0 0 ... 0 0 \n",
- "1 0 ... 0 0 \n",
- "2 0 ... 0 0 \n",
- "3 0 ... 0 0 \n",
- "4 0 ... 0 0 \n",
- "\n",
- " assigned_room_type_K assigned_room_type_L assigned_room_type_P \\\n",
- "0 0 0 0 \n",
- "1 0 0 0 \n",
- "2 0 0 0 \n",
- "3 0 0 0 \n",
- "4 0 0 0 \n",
- "\n",
- " deposit_type_Non Refund deposit_type_Refundable customer_type_Group \\\n",
- "0 0 0 0 \n",
- "1 0 0 0 \n",
- "2 0 0 0 \n",
- "3 0 0 0 \n",
- "4 0 0 0 \n",
- "\n",
- " customer_type_Transient customer_type_Transient-Party \n",
- "0 1 0 \n",
- "1 1 0 \n",
- "2 1 0 \n",
- "3 1 0 \n",
- "4 1 0 \n",
- "\n",
- "[5 rows x 226 columns]"
- ]
- },
- "execution_count": 26,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "df_dummies = pd.get_dummies(df, drop_first=True) # 0-1 encoding for categorical values\n",
- "df_dummies.head()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 27,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "#df_dummies.to_csv('train_dummies.csv', index = False) "
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 28,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/plain": [
- "Int64Index([ 0, 1, 2, 3, 4, 5, 6, 7,\n",
- " 8, 9,\n",
- " ...\n",
- " 119380, 119381, 119382, 119383, 119384, 119385, 119386, 119387,\n",
- " 119388, 119389],\n",
- " dtype='int64', length=118898)"
- ]
- },
- "execution_count": 28,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "df_dummies.axes[0]"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 29,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "data": {
- "text/plain": [
- "Index(['is_canceled', 'lead_time', 'stays_in_week_nights', 'adults',\n",
- " 'children', 'babies', 'is_repeated_guest', 'previous_cancellations',\n",
- " 'booking_changes', 'days_in_waiting_list',\n",
- " ...\n",
- " 'assigned_room_type_H', 'assigned_room_type_I', 'assigned_room_type_K',\n",
- " 'assigned_room_type_L', 'assigned_room_type_P',\n",
- " 'deposit_type_Non Refund', 'deposit_type_Refundable',\n",
- " 'customer_type_Group', 'customer_type_Transient',\n",
- " 'customer_type_Transient-Party'],\n",
- " dtype='object', length=226)"
- ]
- },
- "execution_count": 29,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "df_dummies.axes[1]"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Modellierung und Auswertung",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "# 4. Modellierung und Auswertung"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "Der Datensatz wird mit seinen Dummy-Variablen hochgeladen und in einen Trainings- und einen Testsatz aufgeteilt.\n",
- "Dann wird der Trainings- und Testprozess mit 3 verschiedenen Algorithmen durchgeführt und ausgewertet - Logistische Regression, Entscheidungsbaum,\n",
- "Random Forest.\n",
- "Fur Bewertung, Hyperparameter: \n",
- "Output: überwachtes Lernen, Klassifikation\n",
- "Datenaufteilung: 80% Trainingsdaten, 20% Testdaten.\n",
- "Auswertungsmetriken DecisionTree: Genauigkeit= 0.82, Rückruf= 0.74, Präzision=0.76.\n",
- "Auswertungsmetriken Logistische Regression: Genauigkeit= 0.78, Rückruf= 0.55,\n",
- "Präzision= 0.78.\n",
- "Auswertungsmetriken Random Forest: Genauigkeit= 0.82, Rückruf= 0.57, Genauigkeit= 0,90.\n"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Test- und Trainingsdaten",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 4.1. Test and Train Data"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 30,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [],
- "source": [
- "target = df_dummies['is_canceled'] # feature to be predicted\n",
- "predictors = df_dummies.drop(['is_canceled'], axis = 1) # all other features are used as predictors"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 31,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "data": {
- "text/html": [
- "<div>\n",
- "<style scoped>\n",
- " .dataframe tbody tr th:only-of-type {\n",
- " vertical-align: middle;\n",
- " }\n",
- "\n",
- " .dataframe tbody tr th {\n",
- " vertical-align: top;\n",
- " }\n",
- "\n",
- " .dataframe thead th {\n",
- " text-align: right;\n",
- " }\n",
- "</style>\n",
- "<table border=\"1\" class=\"dataframe\">\n",
- " <thead>\n",
- " <tr style=\"text-align: right;\">\n",
- " <th></th>\n",
- " <th>lead_time</th>\n",
- " <th>stays_in_week_nights</th>\n",
- " <th>adults</th>\n",
- " <th>children</th>\n",
- " <th>babies</th>\n",
- " <th>is_repeated_guest</th>\n",
- " <th>previous_cancellations</th>\n",
- " <th>booking_changes</th>\n",
- " <th>days_in_waiting_list</th>\n",
- " <th>hotel_Resort Hotel</th>\n",
- " <th>...</th>\n",
- " <th>assigned_room_type_H</th>\n",
- " <th>assigned_room_type_I</th>\n",
- " <th>assigned_room_type_K</th>\n",
- " <th>assigned_room_type_L</th>\n",
- " <th>assigned_room_type_P</th>\n",
- " <th>deposit_type_Non Refund</th>\n",
- " <th>deposit_type_Refundable</th>\n",
- " <th>customer_type_Group</th>\n",
- " <th>customer_type_Transient</th>\n",
- " <th>customer_type_Transient-Party</th>\n",
- " </tr>\n",
- " </thead>\n",
- " <tbody>\n",
- " <tr>\n",
- " <td>0</td>\n",
- " <td>342</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>3</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>1</td>\n",
- " <td>737</td>\n",
- " <td>0</td>\n",
- " <td>2</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>4</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>2</td>\n",
- " <td>7</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>3</td>\n",
- " <td>13</td>\n",
- " <td>1</td>\n",
- " <td>1</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " <tr>\n",
- " <td>4</td>\n",
- " <td>14</td>\n",
- " <td>2</td>\n",
- " <td>2</td>\n",
- " <td>0.0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>...</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>0</td>\n",
- " <td>1</td>\n",
- " <td>0</td>\n",
- " </tr>\n",
- " </tbody>\n",
- "</table>\n",
- "<p>5 rows × 225 columns</p>\n",
- "</div>"
- ],
- "text/plain": [
- " lead_time stays_in_week_nights adults children babies \\\n",
- "0 342 0 2 0.0 0 \n",
- "1 737 0 2 0.0 0 \n",
- "2 7 1 1 0.0 0 \n",
- "3 13 1 1 0.0 0 \n",
- "4 14 2 2 0.0 0 \n",
- "\n",
- " is_repeated_guest previous_cancellations booking_changes \\\n",
- "0 0 0 3 \n",
- "1 0 0 4 \n",
- "2 0 0 0 \n",
- "3 0 0 0 \n",
- "4 0 0 0 \n",
- "\n",
- " days_in_waiting_list hotel_Resort Hotel ... assigned_room_type_H \\\n",
- "0 0 1 ... 0 \n",
- "1 0 1 ... 0 \n",
- "2 0 1 ... 0 \n",
- "3 0 1 ... 0 \n",
- "4 0 1 ... 0 \n",
- "\n",
- " assigned_room_type_I assigned_room_type_K assigned_room_type_L \\\n",
- "0 0 0 0 \n",
- "1 0 0 0 \n",
- "2 0 0 0 \n",
- "3 0 0 0 \n",
- "4 0 0 0 \n",
- "\n",
- " assigned_room_type_P deposit_type_Non Refund deposit_type_Refundable \\\n",
- "0 0 0 0 \n",
- "1 0 0 0 \n",
- "2 0 0 0 \n",
- "3 0 0 0 \n",
- "4 0 0 0 \n",
- "\n",
- " customer_type_Group customer_type_Transient customer_type_Transient-Party \n",
- "0 0 1 0 \n",
- "1 0 1 0 \n",
- "2 0 1 0 \n",
- "3 0 1 0 \n",
- "4 0 1 0 \n",
- "\n",
- "[5 rows x 225 columns]"
- ]
- },
- "execution_count": 31,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "predictors.head()"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 32,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [],
- "source": [
- "X_train, X_test, y_train, y_test = train_test_split(predictors, target, test_size=0.2, random_state=123)"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "DecisionTree",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 4.2. DecisionTree"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 33,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "data": {
- "text/plain": [
- "DecisionTreeClassifier()"
- ]
- },
- "execution_count": 33,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "tree = DecisionTreeClassifier()\n",
- "tree.fit(X_train, y_train)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 34,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "12983 2041 2295 6461\n"
- ]
- }
- ],
- "source": [
- "tn, fp, fn, tp = confusion_matrix(y_test, tree.predict(X_test)).ravel() \n",
- "print(tn, fp, fn, tp)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 35,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- " precision recall f1-score support\n",
- "\n",
- " 0 0.97 0.99 0.98 59721\n",
- " 1 0.99 0.94 0.97 35397\n",
- "\n",
- " accuracy 0.98 95118\n",
- " macro avg 0.98 0.97 0.97 95118\n",
- "weighted avg 0.98 0.98 0.98 95118\n",
- "\n"
- ]
- }
- ],
- "source": [
- "print(classification_report(y_train, tree.predict(X_train)))"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 36,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- " precision recall f1-score support\n",
- "\n",
- " 0 0.85 0.86 0.86 15024\n",
- " 1 0.76 0.74 0.75 8756\n",
- "\n",
- " accuracy 0.82 23780\n",
- " macro avg 0.80 0.80 0.80 23780\n",
- "weighted avg 0.82 0.82 0.82 23780\n",
- "\n"
- ]
- }
- ],
- "source": [
- "print(classification_report(y_test, tree.predict(X_test)))"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "paragraph": "Logistik Regression",
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 4.3. Logistic Regression "
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 37,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "name": "stderr",
- "output_type": "stream",
- "text": [
- "C:\\Users\\alexa\\Anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
- "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
- "\n",
- "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
- " https://scikit-learn.org/stable/modules/preprocessing.html\n",
- "Please also refer to the documentation for alternative solver options:\n",
- " https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
- " extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n"
- ]
- },
- {
- "data": {
- "text/plain": [
- "LogisticRegression()"
- ]
- },
- "execution_count": 37,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
- "source": [
- "logreg = LogisticRegression()\n",
- "logreg.fit(X_train, y_train)"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 38,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "[[13634 1390]\n",
- " [ 3870 4886]]\n"
- ]
- }
- ],
- "source": [
- "print(confusion_matrix(y_test, logreg.predict(X_test)))"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 39,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "data": {
- "text/plain": [
- "Text(0.5, 39.5, 'Predicted label')"
- ]
- },
- "execution_count": 39,
- "metadata": {},
- "output_type": "execute_result"
- },
- {
- "data": {
- "image/png": "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\n",
- "text/plain": [
- "<Figure size 720x504 with 2 Axes>"
- ]
- },
- "metadata": {
- "needs_background": "light"
- },
- "output_type": "display_data"
- }
- ],
- "source": [
- "conf_mat = confusion_matrix(y_test, logreg.predict(X_test))\n",
- "df_cm = pd.DataFrame(conf_mat, index=['0','1'], columns=['0', '1'],)\n",
- "fig = plt.figure(figsize=[10,7])\n",
- "heatmap = sns.heatmap(df_cm, annot=True, fmt=\"d\")\n",
- "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=14)\n",
- "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=45, ha='right', fontsize=14)\n",
- "plt.ylabel('True label')\n",
- "plt.xlabel('Predicted label')"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 40,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- " precision recall f1-score support\n",
- "\n",
- " 0 0.78 0.91 0.84 15024\n",
- " 1 0.78 0.56 0.65 8756\n",
- "\n",
- " accuracy 0.78 23780\n",
- " macro avg 0.78 0.73 0.74 23780\n",
- "weighted avg 0.78 0.78 0.77 23780\n",
- "\n"
- ]
- }
- ],
- "source": [
- "print(classification_report(y_test, logreg.predict(X_test)))"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 41,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- " precision recall f1-score support\n",
- "\n",
- " 0 0.78 0.91 0.84 59721\n",
- " 1 0.78 0.56 0.66 35397\n",
- "\n",
- " accuracy 0.78 95118\n",
- " macro avg 0.78 0.74 0.75 95118\n",
- "weighted avg 0.78 0.78 0.77 95118\n",
- "\n"
- ]
- }
- ],
- "source": [
- "print(classification_report(y_train, logreg.predict(X_train)))"
- ]
- },
- {
- "cell_type": "markdown",
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": []
- },
- "source": [
- "## 4.4. Random Forest"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 42,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Max tree depth: 5\n",
- "Train results: precision recall f1-score support\n",
- "\n",
- " 0 0.73 1.00 0.84 59721\n",
- " 1 1.00 0.37 0.54 35397\n",
- "\n",
- " accuracy 0.77 95118\n",
- " macro avg 0.86 0.69 0.69 95118\n",
- "weighted avg 0.83 0.77 0.73 95118\n",
- "\n",
- "Test results: precision recall f1-score support\n",
- "\n",
- " 0 0.73 1.00 0.84 15024\n",
- " 1 1.00 0.36 0.53 8756\n",
- "\n",
- " accuracy 0.76 23780\n",
- " macro avg 0.86 0.68 0.69 23780\n",
- "weighted avg 0.83 0.76 0.73 23780\n",
- "\n",
- "Max tree depth: 10\n",
- "Train results: precision recall f1-score support\n",
- "\n",
- " 0 0.75 0.98 0.85 59721\n",
- " 1 0.94 0.45 0.61 35397\n",
- "\n",
- " accuracy 0.78 95118\n",
- " macro avg 0.85 0.71 0.73 95118\n",
- "weighted avg 0.82 0.78 0.76 95118\n",
- "\n",
- "Test results: precision recall f1-score support\n",
- "\n",
- " 0 0.75 0.98 0.85 15024\n",
- " 1 0.94 0.44 0.59 8756\n",
- "\n",
- " accuracy 0.78 23780\n",
- " macro avg 0.84 0.71 0.72 23780\n",
- "weighted avg 0.82 0.78 0.76 23780\n",
- "\n",
- "Max tree depth: 20\n",
- "Train results: precision recall f1-score support\n",
- "\n",
- " 0 0.81 0.97 0.88 59721\n",
- " 1 0.92 0.62 0.74 35397\n",
- "\n",
- " accuracy 0.84 95118\n",
- " macro avg 0.87 0.79 0.81 95118\n",
- "weighted avg 0.85 0.84 0.83 95118\n",
- "\n",
- "Test results: precision recall f1-score support\n",
- "\n",
- " 0 0.80 0.96 0.87 15024\n",
- " 1 0.90 0.58 0.70 8756\n",
- "\n",
- " accuracy 0.82 23780\n",
- " macro avg 0.85 0.77 0.79 23780\n",
- "weighted avg 0.83 0.82 0.81 23780\n",
- "\n"
- ]
- }
- ],
- "source": [
- "tree_depth = [5, 10, 20]\n",
- "for i in tree_depth:\n",
- " rf = RandomForestClassifier(max_depth=i)\n",
- " rf.fit(X_train, y_train)\n",
- " print('Max tree depth: ', i)\n",
- " print('Train results: ', classification_report(y_train, rf.predict(X_train)))\n",
- " print('Test results: ',classification_report(y_test, rf.predict(X_test)))"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": 43,
- "metadata": {
- "editable": true,
- "include": true,
- "slideshow": {
- "slide_type": ""
- },
- "tags": [
- "active_ipynb"
- ]
- },
- "outputs": [
- {
- "name": "stdout",
- "output_type": "stream",
- "text": [
- "Max tree depth: 20\n",
- "Train results: precision recall f1-score support\n",
- "\n",
- " 0 0.80 0.98 0.88 59721\n",
- " 1 0.94 0.60 0.73 35397\n",
- "\n",
- " accuracy 0.84 95118\n",
- " macro avg 0.87 0.79 0.81 95118\n",
- "weighted avg 0.85 0.84 0.83 95118\n",
- "\n",
- "Test results: precision recall f1-score support\n",
- "\n",
- " 0 0.79 0.97 0.87 15024\n",
- " 1 0.91 0.56 0.69 8756\n",
- "\n",
- " accuracy 0.82 23780\n",
- " macro avg 0.85 0.76 0.78 23780\n",
- "weighted avg 0.83 0.82 0.80 23780\n",
- "\n"
- ]
- }
- ],
- "source": [
- "rf = RandomForestClassifier(max_depth=20)\n",
- "rf.fit(X_train, y_train)\n",
- "print('Max tree depth: ', i)\n",
- "print('Train results: ', classification_report(y_train, rf.predict(X_train)))\n",
- "print('Test results: ',classification_report(y_test, rf.predict(X_test)))"
- ]
- }
- ],
- "metadata": {
- "branche": "Tourismus",
- "category": "Hotel",
- "dataSource": "https://www.kaggle.com/jessemostipak/hotel-booking-demand",
- "funktion": "Marketforschung",
- "kernelspec": {
- "display_name": "Python 3 (ipykernel)",
- "language": "python",
- "name": "python3"
- },
- "language_info": {
- "codemirror_mode": {
- "name": "ipython",
- "version": 3
- },
- "file_extension": ".py",
- "mimetype": "text/x-python",
- "name": "python",
- "nbconvert_exporter": "python",
- "pygments_lexer": "ipython3",
- "version": "3.12.3"
- },
- "repoLink": "https://github.com/AlexRossmann/machine-learning-services/tree/main/Marketing/Generation%20of%20Individual%20Playlists",
- "skipNotebookInDeployment": false,
- "teaser": "Ein wichtiges Ziel für jedes Unternehmen liegt in der Kunden Zufriedenheit und Gewinn maximierung. Problematisch ist wenn die Kunde Ihre Buchungen plotzlich stonieren ohne vorkentnisse. Diese kann nicht nur zu Gewinn velust und negative Bewertung das Hotel fuhren sondern auch zu wirtschaft unsicherheit. Abschätzung des Kundenverhaltens in vorraus bei Buchungsstornierungen für Planung von Kapazität ist notwending.. ",
- "title": "Vorherzusagen von Buchungstornierungen"
- },
- "nbformat": 4,
- "nbformat_minor": 4
-}
--
GitLab