Xuan Kan

Department of Computer Science, Emory University

The one in pink is me :)

N410, 201 Dowman Drive

Atlanta, Georgia 30322

My name is Xuan Kan (阚璇). I obtained my bachelor’s degree in Software Engineering at Tongji University, Shanghai, advised by Prof. Lin Zhang. During my undergraduate study, my research focused on pervasive and trustworthy systems design and implementation using machine learning methods supervised by Dr. Xiaoxuan Lu at University of Oxford. Besides, I interned at Smart City Group of SenseTime, working on improving neural network efficiency with Neural Architecture Search.

Currently, I am pursuing my Ph.D. degree in computer science at Emory University under the joint supervision of Prof. Carl Yang and Prof. Ying Guo. My research goal aims to design more efficient and interpretable machine learning algorithms for fMRI data, which can assist in neurobiological findings and mental disease diagnosis.

news

Nov 19, 2022 Our work entitled Neighborhood-Regularized Self-Training for Label-Efficient Learning accepted to AAAI 2023. Congrats to Ran and Yue!
Oct 21, 2022 Our benchmark paper “BrainGB: A Benchmark for Brain Network Analysis with Graph Neural Networks” has been accepted to IEEE TMI (Journal’s impact factor: 10.048).
Oct 17, 2022 Start my part-time researcher internship on Google Federated Assistant Team mentored by Yonghui (Yohu) Xiao and Tien-Ju Yang!
Sep 14, 2022 One paper entitled Brain Network Transformer has been accepted for NeurIPS 2022. The paper and code are available here.
Jun 2, 2022 Our work on cross-data brain networks got accepted to KDD Health Day.

selected publications

  1. NeurIPS
    Brain Network Transformer (Spotlight)
    Kan, Xuan, Dai, Wei, Cui, Hejie, Zhang, Zilong, Guo, Ying, and Yang, Carl
    Proceedings of the Conference on Neural Information Processing Systems 2022
    Also Presented in Workshop ICML-IMLH 2022 (Oral, no proceedings)
    arXiv Code
  2. MIDL
    FBNetGen: Task-aware GNN-based fMRI Analysis via Functional Brain Network Generation (Oral)
    Kan, Xuan, Cui, Hejie, Lukemire, Joshua, Guo, Ying, and Yang, Carl
    Medical Imaging with Deep Learning 2022
    Also Presented in Workshop ICML-IMLH 2021 (no proceedings)
    arXiv Code
  3. ECML-PKDD
    Zero-shot Scene Graph Relation Prediction through Commonsense Knowledge Integration
    Kan, Xuan, Cui, Hejie, and Yang, Carl
    The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2021
    arXiv Code Poster