Department of Physics, Graduate School of Science
Research Introduction
My major is statistical physics, especially information statistical mechanics, which covers a huge number of degrees of freedom (particles, information, ...).
I am studying statistical physics of machine learning in deep neural networks (deep learning).
Deep learning is recognized as very useful in various situations such as image recognition and machine translation.
However, the mechanism of learning remains unclear.
In particular, why is it possible to learn with a much larger number of parameters than the number of data? (Over-parameterized ). This is a big mystery.
The purpose of my research is to answer this mystery using statistical physics methods.
I believe that solving this fundamental problem is also important for advancing the application of machine learning in society.
Self-introduction(Hobbies)
Anime
I like the voice actress Maaya Uchida and have been attending live shows and events since around 2016.
・Games
I like FPS games. Recently I have been playing "VALORANT".