Contrastive Explanations with Local Foil Trees
This work addresses the problem of feature sparsity in explanations for users of interpretable AI, though it appears incremental as it builds on existing contrastive approaches.
The paper tackles the challenge of high-dimensional feature spaces in interpretable machine learning by proposing a method that uses contrastive explanations to reduce the number of features, focusing on those critical for distinguishing between different outputs, and demonstrates it on three benchmark classification tasks.
Recent advances in interpretable Machine Learning (iML) and eXplainable AI (XAI) construct explanations based on the importance of features in classification tasks. However, in a high-dimensional feature space this approach may become unfeasible without restraining the set of important features. We propose to utilize the human tendency to ask questions like "Why this output (the fact) instead of that output (the foil)?" to reduce the number of features to those that play a main role in the asked contrast. Our proposed method utilizes locally trained one-versus-all decision trees to identify the disjoint set of rules that causes the tree to classify data points as the foil and not as the fact. In this study we illustrate this approach on three benchmark classification tasks.