Mixing Art into the Science of Model Explainability

Overview on Explainable Boosting Machine and an approach for converting ML explanation to more human-friendly explanation.

Fig.1 — A lego figure on my desk, Image by the author.

1. Science of ML explainability

1.1 The Interpretability vs Accuracy Trade-off

In traditional tabular machine learning approaches, Data scientists often deal with the trade-off b/w interpretability and accuracy.

Fig.2: Interpretability/Intelligibility and Accuracy Tradeoff, Image by Author

As shown in the chart above, we can see that Glass-Box models like Logistic Regression, Naive …

art data science explainability explainable ai machine learning science tabular data technology