CV
Education
- B.A. in Economics, Rikkyo University, 2011
- M.S. in Economics, Norwegian School of Economics, 2013
Work experience
CyberAgent, Inc. — Tokyo, Japan
Principal Data Scientist
Apr 2022 – Present
Defined and institutionalized the role of data scientists, building training programs that enabled business units to incorporate causal inference into core decision-making processes
Built and led data science teams in business units where data science practices were not yet established, transforming decision-making from intuition-driven to experiment- and data-driven approaches
Decision Division Leader / Economic Research Scientist
2016 - Present
Founded and led the AILab, establishing an R&D function dedicated to improving organizational decision-making through AI, economics, and causal inference
Built and scaled a 15+ PhD-level organization across 3 specialized teams, integrating machine learning, reinforcement learning, economics, and causal inference into a unified decision intelligence capability
Developed domain-specific team leaders and organizational structures to enable scalable, high-impact research aligned with business needs
Collaborated with business executives to define and prioritize decision-critical problems, aligning R&D investments with strategic objectives
Launched and led cross-functional research initiatives with business data science teams, translating operational challenges into deployable analytical solutions
Data Scientist
2013 – 2016
Introduced machine learning into ad tech platforms (DSP, SSP, DMP), including model development and performance evaluation in production environments
Leveraged product data to uncover growth opportunities, informing product development and revenue optimization decisions, resulting in a multi-million USD impact
Mentored junior data scientists, fostering both technical expertise and the ability to connect data analysis with business impact
Data Analyst
2011 – 2013
Analyzed client marketing data within an advertising agency setting, informing budget allocation and creative optimization decisions
Evaluated the effectiveness of advertising campaigns, including large-scale offline (event-based) marketing initiatives
Conducted joint research projects with Japanese universities, applying time-series models (VAR) to measure advertising impact and presenting findings to clients
Publications
2025
Beyond the Average: Distributional Causal Inference under Imperfect Compliance
Undral Byambadalai, Tomu Hirata, Tatsushi Oka, Shota Yasui
Neural Information Processing Systems(NeurIPS) 2025
[arxiv]
On Efficient Estimation of Distributional Treatment Effects under Covariate-Adaptive Randomization
Undral Byambadalai, Tomu Hirata, Tatsushi Oka, Shota Yasui
International Conference on Machine Learning(ICML) 2025
[arxiv]
Regression Adjustment for Estimating Distributional Treatment Effects in Randomized Controlled Trials
Tatsushi Oka, Shota Yasui, Yuta Hayakawa, Undral Byambadalai Econometric Reviews [SSRN]
2024
Estimating Distributional Treatment Effects in Randomized Experiments: Machine Learning for Variance Reduction
Undral Byambadalai, Tatsushi Oka, Shota Yasui
International Conference on Machine Learning(ICML) 2024
[paper]
2022
Delayed Feedback Modeling with a Time Window Assumption
Shota Yasui, Masahiro Kato
ACM SIGKDD Conference on Knowledge Discovery and Data Mining(KDD) 2022
[arxiv]
Learning Causal Relationships from Conditional Moment Restrictions by Importance Weighting
Masahiro Kato, Masaaki Imaizumi, Kenichiro McAlinn, Shota Yasui, Haruo Kakehi
International Conference on Learning Representations(ICLR) 2022
[arxiv]
Evaluating the impact of COVID-19 on ex-vessel prices using time-series analysis
Keita Abe, Gakushi Ishimura, Shinya Baba, Shota Yasui, Kosuke Nakamura
Fisheries Science, 1-12
[URL]
2021
The Adaptive Doubly Robust Estimator for Policy Evaluation in Adaptive Experiments and a Paradox Concerning Logging Policy
Masahiro Kato, Kenichiro McAlinn and Shota Yasui
Neural Information Processing Systems(NeurIPS) 2021
[arxiv]
Debiased Off-Policy Evaluation for Recommendation Systems
Yusuke Narita, Shota Yasui and Kohei Yata
ACM Conference on Recommender Systems(RecSys) 2021
[arxiv]
2020
Off-Policy Evaluation and Learning for External Validity under a Covariate Shift
Kato, Masahiro, Masatoshi Uehara, and Shota Yasui
Neural Information Processing Systems(NeurIPS) 2020
[arxiv]
Counterfactual Cross-Validation: Stable Model Selection Procedure for Causal Inference Models
Saito, Yuta, and Shota Yasui
International Conference on Machine Learning(ICML) 2020,
[arxiv]
Dual Learning Algorithm for Delayed Conversions
Saito, Yuta, Gota Morishita, and Shota Yasui
International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) 2020
[arxiv]
A Feedback Shift Correction in Predicting Conversion Rates under Delayed Feedback
Yasui, Shota, Gota Morishita, Komei Fujita, Masashi Shibata
The Web Conference (WWW) 2020
[arxiv]
Fatigue-Aware Ad Creative Selection
D Moriwaki, K Fujita, S Yasui, T Hoshino
WSDM Workshop on State-based User Modelling (SUM) 2020
[arxiv]
2019
Efficient counterfactual learning from bandit feedback
Narita, Yusuke, Shota Yasui, and Kohei Yata
AAAI Conference on Artificial Intelligence(AAAI) 2019
[arxiv]
Dual Learning Algorithm for Delayed Feedback in Display Advertising
Saito, Yuta, Gota Morishita, and Shota Yasui
NeurIPS Workshop on Causal Machine Learning 2019
[arxiv]
Counterfactual Cross-Validation
Saito, Yuta, and Shota Yasui
RecSys Workshop on Reinforcement and Robust Estimators for Recommendation (REVEAL) 2019
Book
効果検証入門～正しい比較のための因果推論／計量経済学の基礎
安井翔太 著，株式会社ホクソエム 監修.
技術評論社, 2020-01-18
[Amazon-JP]
施策デザインのための機械学習入門〜データ分析技術のビジネス活用における正しい考え方
齋藤優太，安井翔太 著，株式会社ホクソエム 監修.
技術評論社, 2021-08-04
[Amazon-JP]
Pythonで学ぶ効果検証入門
安井 翔太 監修、伊藤 寛武 著、金子 雄祐 著 オーム社, 2024-05-21
[Amazon-JP]
因果推論入門〜ミックステープ：基礎から現代的アプローチまで
Scott Cunningham (著), 加藤 真大 (翻訳), 河中 祥吾 (翻訳), 白木 紀行 (翻訳), 冨田 燿志 (翻訳), 早川 裕太 (翻訳), 兵頭 亮介 (翻訳), 藤田 光明 (翻訳), 邊土名 朝飛 (翻訳), 森脇 大輔 (翻訳), 安井 翔太 (翻訳)
技術評論社, 2023-05-09
[Amazon-JP]
「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
安井 翔太 (著), 伊藤 寛武 (著), 金子 雄祐 (著)
日経BP 日本経済新聞出版, 2026-06-19
[Amazon-JP]