Dynamic Intepretation of Random Forests Predictions

Notes on understanding why Random Forests makes its decisions.

Understanding Random Forests

 

A good and visual explanation of how Random Forests works.

Model Feature Importances

Feature importances can be taken from Scikit-learn and Spark MLLib implementations after training.

However, this explains features as a whole based on the training dataset. i.e. We are still lacking visibility on an individual prediction.

Different methods of Explaining 

A good overview of ways to explain a random forests model.

Visual explanation for each prediction 

This library does the job.

https://pypi.org/project/treeinterpreter/