[10p-N106-1~9] FS.1 Focused Session "AI Electronics"

1:30 PM - 2:00 PM JST(4:30 AM - 5:00 AM UTC)

[10p-N106-1] [The 23rd Molecular Electronics and Bioelectronics Division Encouragement Award Speech] Physical Reservoir Computing with cultured modular neuronal network

〇Takuma Sumi¹, Hideaki Yamamoto^1,2, Yuichi Katori³, Koki Ito², Satoshi Moriya², Tomohiro Konno⁴, Shigeo Sato², Ayumi Hirano-Iwata^1,2 (1.AIMR Tohoku Univ., 2.RIEC Tohoku Univ., 3.Future Univ. Hakodate, 4.GSPS Tohoku Univ.)

Keywords:

reservoir computing,physical reservoir computing,cultured neuronal network

The complex dynamics generated by neuronal networks serve as a resource for information processing in the brain. In this study, we utilized the dynamics of rat cortical neuronal cultures in a reservoir-computing framework. By introducing into the cultured networks a modular architecture that mimics the brain’s network, we demonstrated that such modularity improves the classification of input information.