Machine Learning Service Repository

Goal

The goal of this repository is to help businesses to figure out the value of applying Machine Learning (ML) in their business activity's. With the ML Services documented in this repository you can learn how ML used to create value. In the accompanying notebook and the attached data you can learn by example how the data is explored, how relevant conclusion are drawn from the data visualizations and how a ML model is created, trained and deployed.

Introduction

This repository contains 25 publicly available use cases that can be used as a machine learning service. These use cases include data frames from different application areas such as medicine, marketing, IoT, etc. The models were all implemented using the Python programming language and are stored in this repository as Jupyter Notebooks. The repository was structured according to application areas or industries. The services listed below can be found in the respective folders.
All data files are stored in google cloud storage in this bucket

see List of Use Cases, for a structured list of all examples in this repository
see Usage for an explanation on how to use this repository and how to run the notebooks.

one example Use Case has been deployed by creating a REST API opn top of the tensorflow model. You may look at the code on the repository, check out the API docs or try out the frontend to predict the quality of your red Wine.

List of Use Cases

CRM:

1. Customer Churn Prediction
2. Customer Satisfaction Airlines
3. Increase customer satisfaction
4. Sentiment analysis on amazon reviews

Automotive:

5. Improvement of components for autonomous motor vehicles

Warehouse:

6. Classification of clothing through images

Success Prediction:

7. Prediction of Successful or Failed Startups
8. Prediction of Successfully Financed Projects

Online Retail:

9. Size prediction for online fashion retailer

Rating:

10. Accommodation rating
11. What Quality does the Red wine have
12. Digital Valuation of Real Estate

Forecast:

13. Forecast of required vehicles in the city center
14. Sales Forecast for retail store

Health:

15. Risk prediction of heart disease

Insurance:

16. Predicting mental illness for health insurance
17. Prediction Interest for car insurance
18. Insurance Fraud detection

Marketing:

19. Generation of Individual Playlists
20. Predicting clicks on online advertising by Facebook

Tourism:

21. Prediction cancellation of hotel bookings
22. Flight delay prediction

Agriculture:

23. Analysis of the movement and activity of free-ranging cattle

Maintenance:

24. Prediction of IoT system failures based on sensor data

Usage

There are two ways you can run the python notebook yourself, as described below.
Or you click the google colab button in the README, to run the notebook in the cloud (for free).
example:

Setup python virtual environment

To setup the virtual environment download python 3.8, and run the following commands. We recommend using python 3.8, as this release is the most reliable with the tensorflow module.
All notebooks in this repo have been verified to run with python 3.8 and the dependencies listed in requirements.txt

py -3.8 -m venv venv
venv\Scripts\Activate.ps1
pip install --upgrade pip
pip install -r requirements.txt

or (in VSCode) drag and drop the setupPython.ps1 script into your Terminal and press Enter

Then open the notebooks (and README ) inside VSCode
or run

jupyter notebook

in the same terminal.