Keio Univ. Yagami-campus, Building 14, Room 631 A/B
3-14-1 Kouhoku-ku, Hiyoshi, Yokohama 223-8522, JAPAN
Speaker: Akiyoshi Sannai
Date: 23 October 2019
Time: 13:30 - 14:30 (+ 30min for discussion)
Place: Keio University, Yagami campus, Bldg. 14th, Room 631A/B
Title: On graph neural networks
Abstract: Graph neural networks(GNN) are a kind of deep neural networks to treat graphs as input.
Several researchers defined (GNN) by the language of graph, but these are not enough.
We give a way of thinking to treat graph functions as group invariant functions and explain the correspondence between algebraic functions over graphs and invariant subring of a polynomial ring by suitable action.
If you are interested, feel free to join us.
Public events of RIKEN Center for Advanced Intelligence Project (AIP)
Join community