Leveraging Association Rules for Better Predictions and Better Explanations

TL;DR

The paper addresses enhancing predictive accuracy and explanations for tree-based classifiers by integrating symbolic knowledge through association rules mined from a binarized data representation of the model. It translates data into boolean features $X$ under a domain theory $Th$, mines 100% confident classification rules $A_c$ to rectify the forest $F$ into $F^{A_c}$, and computes abductive explanations via UP-majoritary reasoning with an extended theory $Th_e$ that incorporates non-classification rules. The approach yields improvements in predictive performance on most datasets and can substantially reduce the size of abductive explanations, while maintaining tractable explanation generation. The work also discusses controllability, computational efficiency, and potential extensions toward stronger logical entailment through theory compilation.

Abstract

We present a new approach to classification that combines data and knowledge. In this approach, data mining is used to derive association rules (possibly with negations) from data. Those rules are leveraged to increase the predictive performance of tree-based models (decision trees and random forests) used for a classification task. They are also used to improve the corresponding explanation task through the generation of abductive explanations that are more general than those derivable without taking such rules into account. Experiments show that for the two tree-based models under consideration, benefits can be offered by the approach in terms of predictive performance and in terms of explanation sizes.