Weihua Hu (胡 緯華)

400px 

I am a first-year Computer Science Ph.D. student at Stanford.

I received a B.E. in Mathematical Engineering in 2016, and an M.S. in Computer Science in 2018, both from the University of Tokyo, where I worked with Prof. Masashi Sugiyama on machine learning and Prof. Hirosuke Yamamoto on information theory. I also worked with Prof. Jun'ichi Tsujii and Prof. Hideki Mima on natural language processing.

[CV] [Google Scholar]

Research Interests

Reliable machine learning in the wild

Building machine learning systems that can be deployed reliably in the real world.

Unsupervised and semi-supervised deep learning

Leveraging large amounts of data to learn useful representations.

Machine learning for graph-structured data

Developing machine learning methods that can better handle graph-structured data.

Publications

2019

  1. Keyulu Xu*, Weihua Hu*, Jure Leskovec, Stefanie Jegelka.
    How Powerful are Graph Neural Networks?
    International Conference on Learning Representations (ICLR), 2019. (oral)
    [OpenReview] [arXiv]

2018

  1. Weihua Hu, Gang Niu, Issei Sato, Masashi Sugiyama.
    Does Distributionally Robust Supervised Learning Give Robust Classifiers?
    International Conference on Machine Learning (ICML2018), pp.2034-2042, 2018.
    [arXiv]

  2. Bo Han, Quanming Yao, Xingrui Yu, Gang Niu, Miao Xu, Weihua Hu, Ivor Tsang, Masashi Sugiyama.
    Co-teaching: Robust training of deep neural networks with noisy labels.
    Neural Information Processing Systems (NeurIPS2018), to appear, 2018.
    [arXiv]

2017

  1. Weihua Hu, Takeru Miyato, Seiya Tokui, Eiichi Matsumoto, Masashi Sugiyama.
    Learning Discrete Representations via Information Maximizing Self Augmented Training.
    International Conference on Machine Learning (ICML2017), pp.1558-1567, 2017.
    [arXiv][code][talk]

  2. Weihua Hu, Hirosuke Yamamoto, Junya Honda.
    Worst-case Redundancy of Optimal Binary AIFV Codes and their Extended Codes.
    IEEE Transactions on Information Theory, vol.63, no.8, pp.5074-5086, August 2017.
    [arXiv]

  3. Takashi Ishida, Gang Niu, Weihua Hu, Masashi Sugiyama.
    Learning from Complementary Labels.
    Neural Information Processing Systems (NeurIPS2017), pp.5644-5654, 2017.
    [arXiv]

2016

  1. Weihua Hu, Hirosuke Yamamoto, Junya Honda.
    Tight Upper Bounds on the Redundancy of Optimal Binary AIFV Codes.
    IEEE International Symposium on Information Theory (ISIT2016), pp.6–10, 2016.
    [paper][slide]

  2. Weihua Hu, Jun'ichi Tsujii.
    A Latent Concept Topic Model for Robust Topic Inference Using Word Embeddings.
    The annual meeting of the Association for Computational Linguistics (ACL2016), pp.380–386, 2016.
    [paper][poster][code]

Awards

Work Experiences

Contact

Email: weihuahu [at] stanford.edu
URL: http://web.stanford.edu/~weihuahu/
Github: https://github.com/weihua916/