If you want to explain the output of your machine learning model, use SHAP.
In the code above, I use SHAP’s summary plot to visualize the overall impact of features in a DataFrame.
My full article about SHAP.
Link to SHAP.