Hoang NT

PhD Candidate at Tokyo Tech and visiting researcher at RIKEN AIP.

Room 504, W8E Ookayama Campus

Tokyo Institute of Technology

Meguro-ku, Tokyo, Japan

I am a PhD candidate at the School of Computing, Tokyo Institute of Technology. My research interests include machine learning for graphs and kernel methods. I am extremely fortunate to work with Dr. Takanori Maehara at the RIKEN Center for Advanced Intelligence Project since 2019 on graph embedding theory. Currently, I am a visiting researcher at the Functional Analytic Unit of RIKEN AIP, hosted by Dr. Ha Quang Minh.

Download my resume.

My research goal is to understand the limitation and prospect of graph-structured data learning by developing analysis tools from graphon theory and spectral graph theory. All information and opinion on this website is my own.

I enjoy long distance cycling in Japan.

news

Jun 30, 2021 I am an organizer for NeurIPS Meetup Japan 2021.
Jun 25, 2021 I cycled from Osaka to Tokyo in 2 days.
Jan 27, 2021 I gave a talk on Graph Homomorphism Convolution at RIKEN AIP.

selected publications

  1. ICML
    Graph Homomorphism Convolution
    NT, Hoang, and Maehara, Takanori
    In 37th International Conference on Machine Learning 2020
    HTML PDF
  2. ICPR
    Revisiting Graph Neural Networks: Graph Filtering Perspective
    NT, Hoang, Maehara, Takanori, and Murata, Tsuyoshi
    In 25th International Conference on Pattern Recognition (ICPR) 2021
    HTML
  3. arXiv
    A Simple Proof of the Universality of Invariant/Equivariant Graph Neural Networks
    Takanori, Maehara, and NT, Hoang
    2019
    PDF