Takahisa TODA (戸田 貴久)

There are growing concerns regarding the fairness of Machine Learning (ML) algorithms. Individual fairness testing is introduced to address the fairness concerns, and it aims to detect discriminatory instances which exhibit unfairness in a given classifier from its input space. XGBoost is one of the most prominent ML algorithms in recent years. In this study, we propose an individual fairness testing method for XGBoost classifier, leveraging the formal verification technique. To evaluate our method, we build XGBoost classifiers on three real-world datasets, and conduct individual fairness testing against them. Through the evaluation, we observe that our method can correctly detect discriminatory instances in XGBoost classifiers within an acceptable running time. Among all testing tasks, the longest running time for detecting 100 discriminatory instances is 2656.4 seconds.

The Werewolf game is a game with incomplete information in which knowledge about the other players, such as roles assigned to players, is unknown and the game continues with debate about uncertain information or actions taken by players with uncertain beliefs until a win-condition is achieved. Aiming to develop an artificial intelligence that plays the Werewolf game, we consider a logical model for agents with uncertain beliefs based on BDI logic, and present a transformation of logical formulas from BDI logic to first-order logic so as to perform abductive inference on the private information of other players. We also discuss several issues of this approach to be resolved.