Naoki Yoshida (Ph.D.)
Title
Team Director
Members
-
Team directorNaoki Yoshida
Introduction
We develop AI technologies that enable rapid analysis of massive datasets delivered from large telescopes and from physics experiments. We use modern machine learning methods and generative models to extract important signals hidden in noisy data, aiming to discover new celestial objects and phenomena, which may eventually lead to discovery of novel physical laws. We develop an innovative statistical model called “emulator”, which replaces costly nonlinear simulations with ultra-fast evaluation of physical quantities for statistical analysis.
Main Research Field
Mathematical & Physical Sciences
Research Field
Informatics / Astrophysics / Data science / Computational physics
Research Subjects
Numerical simulation
Baysean inference
Generative model
Baysean inference
Generative model
RIKEN Website URL
