/MLPipeline-TelcoCustomerChurn

ML Pipeline for Telco Customer Churn

Primary LanguageJupyter Notebook

Machine Learning DevOps - ML Pipeline - Telco Customer Churn Prediction

This repo if for Machine Learning DevOps project based on Telco Customer Churn Prediction

This pipeline consists of following stages:

  1. Getting data from the data source
  2. Data Cleaning
  3. Train, val, test set data split
  4. Model training

Current Stage: Built Reproducible Model Workflow

Weights and Biases project screenshot:

Screenshot

Metrics:

Trained on Random Forest Model available on scikit-learn package. Achieved accuracy random forest model is 79% and this is not what maximum can be done with this dataset and random forest model.

TODO list:

  1. Build Reproducible Model Workflow DONE ✅
  2. Deploy a Scalable ML Pipeline in Production
  3. ML Model Scoring and Monitoring