Kakei Yamamoto

PhD student

Massachusetts Institute of Technology

Biography

I am a first-year Ph.D. student at Massachusetts Institute of Technology (MIT) in the Department of Electrical Engineering and Computer Science (EECS) under the supervision of Prof. Martin Wainwright. I am also affiliated to the Laboratory for Information and Decision Systems (LIDS) and the Institute for Data, Systems, and Society (IDSS). My research interests lie at the intersection of statistics, machine learning, and optimization, recently focusing on the theory of conditional diffusion models.

Prior to MIT, I graduated in 2023 from the University of Tokyo (UTokyo) with a Bachelors of Engineering in Mathematical Informatics where I performed research on optimization for reinforcement learning in the Department of Applied Mathematics under Prof. Taiji Suzuki.

Interests

Machine learning
Statistical inference
Diffusion model

Education

PhD in Computer Science, 2028 (expected)
Massachusetts Institute of Technology
BEng in Mathematical Informatics, 2023
The University of Tokyo

Recent Publications

Quickly discover relevant content by filtering publications.

Kakei Yamamoto, Yusuke Sakemi, Kazuyuki Aihara (2024). Can Timing-Based Backpropagation Overcome Single-Spike Restrictions in Spiking Neural Networks?. IEEE International Joint Conference on Neural Networks.

Cite

Kakei Yamamoto, Kazusato Oko, Zhuoran Yang, Taiji Suzuki (2024). Mean Field Langevin Actor-Critic: Faster Convergence and Global Optimality beyond Lazy Learning. International Conference on Machine Learning.

Cite

Juno Kim, Kakei Yamamoto, Kazusato Oko, Zhuoran Yang, Taiji Suzuki (2024). Symmetric Mean-field Langevin Dynamics for Distributional Minimax Problems. International Conference on Learning Representations.

Cite

Yusuke Sakemi, Kakei Yamamoto, Takeo Hosomi, Kazuyuki Aihara (2023). Sparse-firing regularization methods for spiking neural networks with time-to-first-spike coding. Scientific Reports.

Cite