DSC internal talk in April

Prof. Kazushi Ikeda gave a lecture in April.

The details are as follows.

====================

Prof. Kazushi Ikeda (Mathematical Informatics Laboratory)

TITLE:
Common diffusion models in artificial intelligence and biological phenomena

ABSTRACT：
Recent theoretical analyses of deep learning reveal that its essence is diffusion and reverse diffusion. Diffusion is also important in biological phenomena since reaction-diffusion models play an important role in pattern generation. This project aims to characterize diffusion/reverse diffusion in both fields and to utilize the knowledge of one field to the other.
Thanks to the support of the DSC Grants for new issue identification activities, we improved the deep learning analysis and found a diffusion process in biological phenomena.