研究業績リスト
ジャーナル論文 - rm_misc: Summary National Conference
大規模言語モデルのコード生成タスクにおけるProperty-Based Testingの活用
公開済 03/2026
研究報告ソフトウェア工学(SE) 研究報告, 2026-SE-222, 14, 1 - 8
ジャーナル論文 - rm_misc: Summary National Conference
初期配置が指定された場合にロジックのみで可解な数独生成問題に対する焼きなまし法に基づく手法の開発
公開済 03/2026
研究報告ゲーム情報学(GI) 研究報告, 2026-GI-57, 14, 1 - 6
ジャーナル論文 - rm_misc: Summary National Conference
公開済 01/2026
人工知能学会研究会資料 人工知能基本問題研究会, 135, 1, 54 - 58
ジャーナル論文 - rm_misc: Summary National Conference
機械学習のブラックボックス公平性テストにおける代理モデルの近似性能の改善
公開済 01/2026
人工知能学会研究会資料 人工知能基本問題研究会, 135, 1, 73 - 78
ジャーナル論文 - rm_misc: Summary National Conference
公開済 01/2026
研究報告アルゴリズム(AL) 研究報告, 2026-AL-206, 17, 1 - 8
ジャーナル論文 - rm_misc: Summary National Conference
公開済 01/2026
人工知能学会研究会資料 人工知能基本問題研究会, 135, 1, 79 - 84
ジャーナル論文 - rm_misc: Summary National Conference
公開済 02/2025
信学技報, 124, 424, 1 - 8
ジャーナル論文 - rm_published_papers: International Conference Proceedings
Toward Individual Fairness Testing with Data Validity
公開済 27/10/2024
Proceedings of the 39th IEEE/ACM International Conference on Automated Software Engineering, 2284 - 2288
ジャーナル論文 - rm_misc: Summary National Conference
Toward Individual Fairness Testing for XGBoost Classifier through Formal Verification
公開済 05/2024
Proceedings of the Annual Conference of JSAI, JSAI2024, 2L6OS19b04 - 2L6OS19b04
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.
ジャーナル論文 - rm_misc: Summary National Conference
公開済 05/2024
Proceedings of the Annual Conference of JSAI, JSAI2024, 4Xin285 - 4Xin285
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.