Interpretable Concept-Based Classification with Shapley Values

P. 90–102.

Ignatov D. I., Kwuida L.

Among the family of rule-based classification models, there are classifiers based on conjunctions of binary attributes. For example, the JSM-method of automatic reasoning (named after John Stuart Mill) was formulated as a classification technique in terms of intents of formal concepts as classification hypotheses. These JSM-hypotheses already represent an interpretable model since the respective conjunctions of attributes can be easily read by decision-makers and thus provide plausible reasons for model prediction. However, from the interpretable machine learning viewpoint, it is advisable to provide decision-makers with importance (or contribution) of individual attributes to the classification of a particular object, which may facilitate explanations by experts in various domains with high-cost errors like medicine or finance. To this end, we use the notion of Shapley value from cooperative game theory, also popular in machine learning. We provide the reader with theoretical results, basic examples, and attribution of JSM-hypotheses by means of Shapley value on real data.

Keywords: Shapley value, индекс Шепли-Шубика, ДСМ-метод автоматического порождения гипотез, JSM method, ДСМ-метод, формальные понятия, Interpretable Machine Learning, formal concepts, explainable artificial intelligence, explainable AI, индекс Шепли, интерпретируемое машинное обучение, объяснимый ИИ, JSM hypotheses, JSM

Publication based on the results of: Development of Mathematical Models and Methods for Recommender Systems and Natural Language Processing (2020)

In book

Ontologies and Concepts in Mind and Machine. 25th International Conference on Conceptual Structures, ICCS 2020

Springer, 2020.