I’m Principal Data Scientist and Senior Researcher at CyberAgent, Inc..

My research lies at the intersection of machine learning, causal inference, and economics. I focus on distributional treatment effects, sampling and selection bias, and randomized experiments, with applications to the evaluation of advertising and recommendation systems.

Research Topics

• Causal Inference
  • Distributional Treatment Effects
  • Randomized Experiments
  • Sampling and Selection Bias
• Machine Learning for Causal Inference
  • Individual Treatment Effect Prediction
  • Uplift Modeling
• Counterfactual Machine Learning
  • Off-Policy Evaluation and Learning
• Advertising and Recommendation Systems
  • Evaluation of Advertising and Recommendations
  • Delayed Feedback
  • Advertising Auctions

Recent Updates

• 2026: Our paper “Where to Cut Ad Load on AVOD: Targeting the Entry Point of Serialized Content on ABEMA” by Shingo Uto and Shota Yasui has been accepted for presentation at CODE@MIT 2026.
• 2026: Our paper “Modeling Covariate Transition for Efficient Estimation of Longitudinal Treatment Effects in Randomized Experiments” by Naoki Chihara, Tatsushi Oka, Yasuko Matsubara, Yasushi Sakurai, and Shota Yasui was published in the Proceedings of ICML 2026. paper, arXiv
• 2026.06.19: Our coauthored book 『「原因」と「結果」を武器にする思考　エビデンスベースで成果を上げる効果検証』 was published. Amazon-JP
• 2026: We released the preprint “Distributional Treatment Effects of Content Promotion: Evidence from an ABEMA Field Experiment” by Shota Yasui, Tatsushi Oka, Undral Byambadalai, and Yuki Oishi. arXiv
• 2025.09.19: Our paper “Beyond the Average: Distributional Causal Inference under Imperfect Compliance” has been accepted to NeurIPS’25
• 2025.05.01: Our paper “On Efficient Estimation of Distributional Treatment Effects under Covariate-Adaptive Randomization” has been accepted to ICML’25

See Publications and CV for the full list of publications and professional experience.