I will be joining Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) in UAE as an Assistant Professor of Machine Learning starting January 2026. My research interests focus on Machine Learning Methodology, with particular emphasis on Multi-armed Bandits, Stochastic Search and Reasoning, Mechanism Design, and Algorithmic Bias.

Prior to joining MBZUAI, I was an Assistant Professor at New York University Stern School of Business. Before that, I was a research associate (post-doctoral researcher) in the University of Tokyo. I currently serve as an area chair for NeurIPS, ICML, and AISTATS, and as an associate editor for Transactions on Machine Learning Research (TMLR).

🎯 For students visiting this webpage during recruiting weeks: Check out my Lightning Talk and Research Overview for a quick overview of my research. Feel free to drop by my office!

I am looking for graduate students as well as visiting researchers. For more details about research opportunities, please see my Join My Research Group page or shoot me an email at junpei@komiyama.info.

Selected Publications with Summary

• Minimax Optimal Algorithms for Fixed-Budget Best Arm Identification (NeurIPS 2022): We characterized the optimal algorithm and performance bound in the fixed-budget best arm identification, which had been open since its modern formulation in 2010.

• Anytime Capacity Expansion in Medical Residency Match by Monte Carlo Tree Search (IJCAI 2022): We applied Monte Carlo tree search for a flexible-capacity mechanism design of matching doctors to hospitals, which is NP-Complete to solve exactly.

• Optimal Regret Analysis of Thompson Sampling in Stochastic Multi-armed Bandit Problem with Multiple Plays (ICML 2015): We show optimality of posterior sampling, a heuristic proposed in 1930s, in the context of online advertisement selection.

Contact

• Email: junpei_at_komiyama.info
• Alternative emails: junpeikomiyama_at_gmail.com

For more details about my research and publications, please see my publications page and CV.