Yuhe Guo

Gaoling AI School, Renmin University of China. Haidian, Beijing.
E-mail: guoyuhe@ruc.edu.cn
My Chinese name means lotus. Feel free to call me Lotus.

prof_pic.jpg

I am an incoming fourth-year Ph.D student in Gaoling School of AI, Renmin University of China (GASI@RUC), advised by Prof. Zhewei Wei. I previously earned my Bachelor’s and Master’s degrees from RUC. My master journey was primarily about graph databases and word embeddings, advised by Professors Xiaoyong Du and Wei Lu.

Since the beginning of my Ph.D. journey, my primary research has focused on Graph Neural Networks(GNNs), particularly Spectral GNNs. In addition, I am interested in other aspects of graph learning, such as Equivariant and Invariant GNNs (IGNs), rewiring methods, and expressive GNNs.

Also, I have been deepening my understanding of the theoretical foundations of GNNs, including signal processing and numerical algebra.

News   that are not so breaking

Aug 15, 2024 I recently wrote an answer on Zhihu (a Chinese version of Quora) , explaining the physical meaning of convolution via an example of continuously tapping a stick and another example involving multiplication (yes, I applied what I just learned from understanding FFT). My understanding of these topics might be quite shallow, but I hope to deepen my understanding as I continue to share.
Aug 14, 2024 My mates and I studied and illustrated the content of Analysis of Boolean Functions. Additionally, it seems like I’ve learned more of the essence of FFT for a first time. Although elementary, these are all wonderful classics, and I welcome discussions!
Aug 10, 2024 My open-source repository PolyFilterPlayground is updated. It might be helpful for you to tune and reproduce the results of polynomial based spectral GNNs.
Aug 01, 2024 I build my personal webpage :)
Jul 25, 2024 I recently wrote an article on Zhihu. It’s about the constructive proof of SVD’s existence, and the optimization perspective it introduces.


Selected Publications

  1. ICML’23
    optbasis.png
    Graph Neural Networks with Learnable and Optimal Polynomial Bases
    Guo Yuhe, and Wei Zhewei
    Proceedings of the 40th International Conference on Machine Learning, 2023
  2. KDD’23
    clenshaw.png
    Clenshaw Graph Neural Networks
    Guo Yuhe, and Wei Zhewei
    Proceedings of the 29th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2023