2019/06/04 by Xavier Renard, Nicolas Woloszko, Renard, Xavier +5 · 6 citations
Computer Science · Decision Sciences · Mathematics · #Artificial intelligence #Black box #Computer science #Decision tree #Domain (mathematical analysis) #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Fidelity #Interpretability #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and Data Classification #Machine learning #Mathematics #Representation (politics) #Stock Market Forecasting Methods #Surrogate model #Tree (set theory) #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.1906.01297
published in arXiv (Cornell University) (Cornell University) · presented at 2019 ICML Workshop on Human in the Loop Learning (HILL 2019), Long Beach, USA
arxiv created 2019/06/04 · openalex publication_date 2019/06/04 · arxiv updated 2019/06/05 · openalex created_date 2022/07/29 · openalex updated_date 2026/08/06
Interpretable surrogates of black-box predictors trained on high-dimensional\ntabular datasets can struggle to generate comprehensible explanations in the\npresence of correlated variables. We propose a model-agnostic interpretable\nsurrogate that provides global and local explanations of black-box classifiers\nto address this issue. We introduce the idea of concepts as intuitive groupings\nof variables that are either defined by a domain expert or automatically\ndiscovered using correlation coefficients. Concepts are embedded in a surrogate\ndecision tree to enhance its comprehensibility. First experiments on FRED-MD, a\nmacroeconomic database with 134 variables, show improvement in\nhuman-interpretability while accuracy and fidelity of the surrogate model are\npreserved.\n