Predict Customer Churn

• Project Predict Customer Churn of ML DevOps Engineer Nanodegree Udacity

Project Description

This project aims to build an ML system that predict banking customer churn based on several attributes. The project follows Software Engineering best practices to enhance an existing ML solution.

Files and data description

The following tree outline the folder structure of the project. In the base folder, churn_library.py and churn_script_logging_and_tests.py contains the main functionality to read and process the data and build and train ML models to predict churn.

The data folder contains raw data in csv format. The images folder contains results of the Exploratory Data Analysis, and models performance analysis. The logs folder contains the output of the testing script. Finally the models folder contains the trained models saved as Pickle file.

├── churn_library.py
├── churn_script_logging_and_tests.py
├── data
│   └── bank_data.csv
├── images
│   ├── eda
│   │   ├── Churn.png
│   │   ├── Customer_Age.png
│   │   ├── Marital_Status.png
│   │   ├── Total_Trans_Ct.png
│   │   └── cols_heatmap.png
│   └── results
│       ├── Logistic_Regression_results.png
│       ├── Random_Forest_results.png
│       ├── feature_importances.png
│       └── roc_curves_results.png
├── logs
│   └── churn_library.log
├── models
│   ├── logistic_model.pkl
│   └── rfc_model.pkl
├── requirements_py3.6.txt
└── requirements_py3.8.txt

Running Files

From the base folder, use the following command to run churn_library.py:

python churn_library.py

To run the testing script use the following command:

python churn_script_logging_and_tests.py