[3E1-GS-2] Machine learning: Explainable AI (1)

[3E1-GS-2-05] Efficient Construction of a Summary Data Structure for Decision Trees

Towards Explainable Artificial Intelligence

〇Hiroki Arimura¹, Kentaro Kanamori¹, Ye Wang ¹ (1. Hokkaido University)

Keywords:explanation in AI, trustable AI, discrete optimization

In this paper, we report our recent progress on summary data structures, called model sketches, for global analysis of prediction models in the context of explainable machine learning. In particular, we explain our recent proogress on model sketches for a subclass of decision trees. We also discuss possible future directions.