When a CBR in Hand is Better than Twins in the Bush

TL;DR

This paper demonstrates that a case-based reasoning (CBR) model, built from an XGBoost regression model, can outperform the original model in both accuracy and interpretability for flight delay prediction, challenging the trade-off notion.

Contribution

It introduces a novel approach to derive a highly accurate and interpretable CBR model from an XGBoost regression model, establishing a new benchmark for accuracy and interpretability.

Findings

CBR model outperforms XGBoost in local prediction accuracy.

CBR provides a clear global explanation via feature importance.

The approach enables effective local explanations using SHAP and LIME.

Abstract

AI methods referred to as interpretable are often discredited as inaccurate by supporters of the existence of a trade-off between interpretability and accuracy. In many problem contexts however this trade-off does not hold. This paper discusses a regression problem context to predict flight take-off delays where the most accurate data regression model was trained via the XGBoost implementation of gradient boosted decision trees. While building an XGB-CBR Twin and converting the XGBoost feature importance into global weights in the CBR model, the resultant CBR model alone provides the most accurate local prediction, maintains the global importance to provide a global explanation of the model, and offers the most interpretable representation for local explanations. This resultant CBR model becomes a benchmark of accuracy and interpretability for this problem context, and hence it is used to evaluate the two additive feature attribute methods SHAP and LIME to explain the XGBoost regression model. The results with respect to local accuracy and feature attribution lead to potentially valuable future work.