Uncovering black box prediction models using explainable artificial intelligence (XAI)

• In Event: Advances in Machine Learning

Brief Overview

This paper applies explainable AI (XAI) techniques, a rapidly developing AI/machine learning research topic, to political datasets. We show how XAI can uncover the "black box" input-output relation of advanced AI/machine learning prediction models.

Authors

• Budrul Ahsan, Philips Japan

• Sota Kato, Tokyo Foundation for Policy Research

• Takafumi Nakanishi, Musashino University

• Ryotaro Okada, Musashino University

• Iku Yoshimoto, The University of Tokyo