Zhanxing Zhu (朱占星)
@ Peking University (北京大学)
Welcome to my homepage!
I am currently an Assistant Professor at School of Mathematical Sciences, Peking University (北京大学) and Beijing Institute of Big Data Research (北京大数据研究院). I am closely affiliated with Deep Learning Lab of Peking University (北京大学深度学习实验室). Previously I obtained my PhD in machine learning from School of Informatics of University of Edinburgh, UK.
My research interests cover methodology/theory of machine learning and artificial intelligence and their applications in various areas.
- Methodology/theory: deep learning, reinforcement learning, scalable optimization and Bayesian inference methods
- Applications: prediction problems in traffic, data-driven medical imaging, generative models and reinforcement learning for computer graphics and network security
Contact: Room 219, Jingyuan 6th Courtyard, Peking University, Beijing, China (北京大学静园6号院219)
Email: zhanxing.zhu AT pku.edu.cn
- Ph.D students: Bing Yu ( co-supervised with Prof. Weinan E)
- MPhil students: Ju Xu, Ke Sun (co-supervised with Prof. Zhouchen Lin), Yuanjin Zhu, Junzhao Zhang
- Undergraduate: Tianyuan Zhang, Jin Ma, Dinghuai Zhang
- Area Chair/Senior PC for AISTATS 2017, AAAI 2019
- Reviewer for JMLR, TPAMI, NeurIPS, ICML, CVPR, ICCV, AISTATS, AAAI, IJCAI, ACML
Publications/Preprints (Check our Github page for reproducibility.)
* indicates equal contribution.
Learning dynamics in deep learning and optimization
- Jingfeng Wu, Wenqing Hu, Haoyi Xiong, Jun Huan and Zhanxing Zhu. The Multiplicative Noise in Stochastic Gradient Descent: Data-Dependent Regularization, Continuous and Discrete Approximation. arXiv pre-print.
- Bing Yu*, Junzhao Zhang* and Zhanxing Zhu. "On the Learning Dynamics of Two-layer Nonlinear Convolutional Neural Networks". arXiv pre-print.
- [ICML 2019] Zhanxing Zhu*, Jingfeng Wu*, Bing Yu, Lei Wu and Jinwen Ma. The Anisotropic Noise in Stochastic Gradient Descent: Its Behavior of Escaping from Minima and Regularization Effects. 36th International Conference on Machine Learning
- [NeurIPS 2018] Rui Luo, Yaodong Yang, Jianhong Wang, Zhanxing Zhu, Jun Wang. Thermostat-assisted continuously-tempered Hamiltonian Monte Carlo for Bayesian learning. 32nd Annual Conference on Neural Information Processing Systems [Code].
- [NeurIPS 2018] Ju Xu, Zhanxing Zhu. Reinforced Continual Learning. 32nd Annual Conference on Neural Information Processing Systems. [Code]
- [IJCAI 2018] Nanyang Ye, Zhanxing Zhu. Stochastic Fractional Hamiltonian Monte Carlo. 27th International Joint Conference on Artificial Intelligence.
- [ICML 2017 W.] Lei Wu, Zhanxing Zhu and Weinan E. Towards Understanding Generalization of Deep Learning: Perspective of Loss Landscapes. 34th International Conference on Machine Learning.
- [NIPS 2017] Nanyang Ye, Zhanxing Zhu, Rafal K. Mantiuk. Langevin Dynamics with Continuous Tempering for Training Deep Neural Networks. 31st Annual Conference on Neural Information Processing Systems.
- [AAAI 2016] Zhanxing Zhu and Amos Storkey. Stochastic Parallel Block Coordinate Descent for Large-scale Saddle Point Problems. Thirtieth AAAI Conference on Artificial Intelligence
- [NIPS 2015 ] Zhanxing Zhu*, Xiaocheng Shang*, Benedict Leimkuhler and Amos Storkey. Covariance-Controlled Adaptive Langevin Thermostat for Large-Scale Bayesian Sampling. 29th Annual Conference on Neural Information Processing Systems(* indicates equal contribution).
- [ECML/PKDD 2015] Zhanxing Zhu and Amos Storkey. Adaptive Stochastic Primal-Dual Coordinate Descent for Separable Saddle Point Problems. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases
Robustness and adversarial learning
- Dinghuai Zhang*, Tianyuan Zhang*, Yiping Lu*, Zhanxing Zhu and Bin Dong. You Only Propagate Once: Accelerating Adversarial Training Using Maximal Principle. arXiv pre-print. [Code]
- Lei Wu, Zhanxing Zhu, and Cheng Tai. Understanding and Enhancing the Transferability of Adversarial Examples. arXiv pre-print.
- [ICML 2019] Tianyuan Zhang, Zhanxing Zhu. Interpreting Adversarial Trained Convolutional Neural Networks. 36th International Conference on Machine Learning. [Code]
- [CVPR 2019] Bing Yu*, Jingfeng Wu*, Jinwen Ma and Zhanxing Zhu. Tangent-Normal Adversarial Regularization for Semi-supervised Learning. The 30th IEEE Conference on Computer Vision and Pattern Recognition. (Oral presentation)
- [NeurIPS 2018] Nanyang Ye, Zhanxing Zhu. Bayesian Adversarial Learning. 32nd Annual Conference on Neural Information Processing Systems. [Code]
- [ECML/PKDD 2019] Ruosi Wan, Mingjun Zhong, Haoyi Xiong and Zhanxing Zhu. Neural Control Variates for Monte Carlo Variance Reduction. The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases. (To Appear).
- [AAAI 2019] Kafeng Wang, Haoyi Xiong, Jiang Bian, Zhanxing Zhu, Chengzhong Xu, Zhishan Guo, Jun Huan. SpHMC: Spectral Hamiltonian Monte Carlo. 33rd AAAI Conference on Artificial Intelligence.
- [ECML/PKDD 2015] Amos Storkey, Zhanxing Zhu and Jinli Hu. Aggregation Under Bias: Renyi Divergence Aggregation and its Implementation via Machine Learning Markets. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases.
- Amos Storkey, Zhanxing Zhu, Jinli Hu. A Continuum from Mixtures to Products: Aggregation under Bias. 31st International Conference on Machine Learning (ICML 2014 Workshop on Divergence Methods for Probabilistic Inference).
ML/DL models and applications
- [Lancet] Xiantong Zou, Xianghai Zhou, Zhanxing Zhu and Linong Ji. Novel subgroups of patients with adult-onset diabetes in Chinese and US populations. Lancet (Diabetes and Endocrinology) 2019.
- [CCS 2018] Guixin Ye, Zhanyong Tang, Dingyi Fang, Zhanxing Zhu, Yansong Feng, Pengfei Xu, Xiaojiang Chen, Zheng Wang. Yet Another Text Captcha Solver: A Generative Adversarial Network Based Approach. 25th ACM Conference on Computer and Communications Security . (Best Paper Award Finalists) [Code]
- [IJCAI 2018] Bing Yu*, Haoteng Yin*, Zhanxing Zhu. Spatio-temporal Graph Convolutional Neural Network: A Deep Learning Framework for Traffic Forecasting. 27th International Joint Conference on Artificial Intelligence . [Code]
- [ISBI 2018] Huizhuo Yuan, Jinzhu Jia and Zhanxing Zhu. SIPID: A Deep Learning Framework for Sinogram Interpolation and Image Denoising in Low-Dose CT Reconstruction. In 2018 IEEE International Symposium on Biomedical Imaging (Oral presentation)[Code]
- [ACL 2017] Bingfeng Luo, Yansong Feng, Zheng Wang, Zhanxing Zhu, Songfang Huang, Rui Yan and Dongyan Zhao. Learning with Noise: Enhance Distantly Supervised Relation Extraction with Dynamic Transition Matrix. 55th Annual Meeting of the Association for Computational Linguistics.
- Francesco Corona, Zhanxing Zhu, Amauri Holanda de Souza Júnior, Michela Mulas, Emanuela Muru, Lorenzo Sassu, Guilherme Barreto, and Roberto Baratti. Supervised Distance Preserving Projections: Applications in the quantitative analysis of diesel fuels and light cycle oils from NIR spectra. Journal of Process Control (2014)
- Zhanxing Zhu, Zhirong Yang and Erkki Oja. Multiplicative Updates for Learning with Stochastic Matrices. In the 18th Conference Scandinavian Conferences on Image Analysis (SCIA 2013) (Oral presentation).
- Zhanxing Zhu, Timo Simila and Francesco Corona. Supervised Distance Preserving Projection. Neural Processing Letters 38(3): 445-463 (2013)
- Zhanxing Zhu, Francesco Corona, Amaury Lendasse, Roberto Baratti and Jose A. Romagnoli. Local linear models for soft-sensor design with application to an industrial deethanizer. 18th World Congress of the International Federation of Automatic Control (IFAC) , Milan, Italy, 2011.
- Zhirong Yang, Zhanxing Zhu and Erkki Oja. Automatic Rank Determination in Projective Nonnegative Matrix Factorization. 9th International Conference on Latent Variable Analysis and Signal Separation (LVA/ICA 2010).
" Only in silence the word, only in dark the light, only in dying life: bright the hawk's flight on the empty sky. "
— Ursula K. Le Guin (A Wizard of Earthsea)
Recently, I wrote a Chinese book with several colleagues on the introduction to data science, which is now available for online shopping. "数据科学导引". Wellcome for all kinds of feedback!
- Ruosi Wan (now research scientist @ Face++)
- Haoteng Yin (now Ph.D student @ Purdue University)
- Jingfeng Wu (now Ph.D student @ John Hopkins University)
- Zizhuo Zhang
- Mengzhang Li