This is a personal project to use explainer-dashboard package.
Note that I got the same key insight from the EDA using a python package called ydata-profiling see my ydata-profiling project here
Key Insight from this EDA: Customer "Complain" is the highest contributor to "Exited" based on SHAP values.
See pdf file for SHAP values
Benefit: This package and explainer automates common exploratory analyses.
I only needed to input the model I wanted to use, which is the Random Forest Classifier Explainer.
Time savings can be allocated to deeper analysis in the future.
Data source: Data are from customer churn data shared by [@mo-chen in YouTube] (https://www.youtube.com/@mo-chen). csv file
Code: Check out details of the code used in Jupyter Notebook here