Hello World!

My name is Xiaoyun Li. I’m a machine learning scientist at LinkedIn, developing algorithms and theory for big data and machine learning. Before joining Linkedin, I was a research scientist at Baidu Research.

Education

• B.S. Financial Mathematics, Shanghai Jiao Tong University, 2015
• M.S. Financial Statistics and Risk Management, Rutgers University, 2017
• Ph.D Statistics, Rutgers University, 2021
  I was fortunate to work with Prof. Ping Li.
  Thesis committee: Prof. Ping Li, Prof. Cun-Hui Zhang, Prof. Min Xu, Prof. William Steiger.

Work Experience

• Teaching Assistant, Department of Statistics, Rutgers University
  – Introduction to Statistics, Advanced Simulation, Data Mining

• Baidu Research Institute

• LinkedIn

Interests

Statistics, Machine Learning, Randomized Algorithms, Information Retrieval, Privacy, Distributed Optimization, Deep Learning, Basketball, Guitar… and anything interesting!

Academic Services

• Conference Reviewer/PC Member: Since 2020, I have served as the reviewer/PC/SPC for NeurIPS, ICML, ICLR, SODA, AISTATS, KDD, WWW, AAAI, IJCAI conference
• Journal Reviewer: IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), Machine Learning, IEEE/CAA Journal of Automatica Sinica

Selected Publications

(* : alphabetical order)

New:

• Xiaoyun Li and Ping Li. Analysis of Error Feedback in Federated Non-convex Optimization with Biased Compression, [Arxiv].
• Ping Li and Xiaoyun Li. OPORP: One Permutation + One Random Projection, [Arxiv].
• Long Feng^*, Tiefeng Jiang^*, Xiaoyun Li^* and Binghui Liu^*. Asymptotic Independence of the Sum and Maximum of Dependent Random Variables with Applications to High-dimensional Tests, Statistica Sinica (2023), [paper].
• Farzin Haddadpour^*, Belhal Karimi^*, Ping Li^* and Xiaoyun Li^*. FedSKETCH: Communication-Efficient and Private Federated Learning via Sketching, [Arxiv].

Dimensionality Reduction & Compression:

• Xiaoyun Li and Ping Li. Generalization Error Analysis of Quantized Compressive Learning, Neural Information Processing Systems (NeurIPS) 2019 (Oral spotlight), [paper].
• Xiaoyun Li and Ping Li. Random Projections with Asymmetric Quantization, Neural Information Processing Systems (NeurIPS) 2019, [paper].
• Xiaoyun Li and Ping Li. Quantization Algorithms for Random Fourier Features, International Conference on Machine Learning (ICML) 2021, [paper].
• Xiaoyun Li and Ping Li. One-Sketch-for-All: Non-linear Random Features from Compressed Linear Measurements, International Conference on Artificial Intelligence and Statistics (AISTATS) 2021, [paper].
• Xiaoyun Li, Chenxi Wu and Ping Li. IVFS: Simple and Efficient Feature Selection for High Dimensional Topology Preservation, AAAI Conference on Artificial Intelligence (AAAI) 2020, [paper].
• Zhiqiang Xu, Dong Li, Weijie Zhao, Xing Shen, Tianbo Huang, Xiaoyun Li, and Ping Li. Agile and Accurate CTR Prediction Model Training for Massive-Scale Online Advertising Systems, ACM International Conference on Management of Data (SIGMOD) 2021, [paper].

Randomized Algorithms, Hashing, Kernel Methods:

• Xiaoyun Li and Ping Li. SignRFF: Sign Random Fourier Features, Neural Information Processing Systems (NeurIPS) 2022, [paper].
• Xiaoyun Li and Ping Li. C-MinHash: Improving Min-wise Hashing with Circulant Permutation, International Conference on Machine Learning (ICML) 2022, [paper].
• Ping Li^*, Xiaoyun Li^* and Cun-Hui Zhang^*. Re-randomized Densification for One Permutation Hashing and Bin-wise Consistent Weighted Sampling, Neural Information Processing Systems (NeurIPS) 2019, [paper].
• Ping Li^*, Xiaoyun Li^*, Gennady Samorodnitsky^* and Weijie Zhao^*. Consistent Sampling Through Extremal Process, The Web Conference (WWW) 2021, [paper].
• Xiaoyun Li and Ping Li. Rejection Sampling for Weighted Jaccard Similarity Revisited, AAAI Conference on Artificial Intelligence (AAAI) 2021, [paper].
• Ping Li^*, Xiaoyun Li^* and Gennady Samorodnitsky^*. P-MinHash Algorithm for Continuous Probability Measures: Theory and Application to Machine Learning, Conference on Information and Knowledge Management (CIKM) 2022, [paper].
• Peng Yang, Xiaoyun Li and Ping Li. Graph-based Adversarial Online Kernel Learning with Adaptive Embedding, International Conference on Data Mining (ICDM) 2021, [paper].
• Xiaoyun Li, Jie Gui and Ping Li. Randomized Kernel Multi-view Discriminant Analysis, European Conference on Artificial Intelligence (ECAI) 2020, [paper].

Distributed/Federated Optimization:

• Xiaoyun Li, Belhal Karimi and Ping Li. On Distributed Adaptive Optimization with Gradient Compression, International Conference on Learning Representations (ICLR) 2022, [paper].
• Xiangyi Chen, Xiaoyun Li and Ping Li. Towards Communication-Efficient Adaptive Gradient Method, ACM-IMS Foundations of Data Science Conference (FODS) 2020, [paper].
• Jun-Kun Wang, Xiaoyun Li, Belhal Karimi and Ping Li. An Optimistic Acceleration of AMSGrad for Nonconvex Optimization, Asian Conference on Machine Learning (ACML) 2021, [paper].
• Weijie Zhao, Xuewu Jiao, Mingqing Hu, Xiaoyun Li, Xiangyu Zhang, and Ping Li. Communication-Efficient Terabyte-Scale Model Training Framework for Online Advertising, IEEE International Conference on Big Data (IEEE BigData), to appear.

Privacy, Causal Inference:

• Chenglin Fan^*, Ping Li^* and Xiaoyun Li^*. Private Graph All-Pairwise-Shortest-Path Distance Release with Improved Error Rate, Neural Information Processing Systems (NeurIPS) 2022, [paper].
• Huang Fang, Xiaoyun Li, Chenglin Fan and Ping Li. Improved Convergence of Differential Private SGD with Gradient Clipping, International Conference on Learning Representations (ICLR) 2023, [paper].
• Xueyan Niu, Xiaoyun Li and Ping Li. Learning Cluster Causal Diagrams: An Information-Theoretic Approach, International Joint Conference on Artificial Intelligence (IJCAI) 2022, [paper].

Computer Vision:

• Tan Yu, Xiaoyun Li and Ping Li. Fast and Compact Bilinear Pooling by Shifted Random Maclaurin, AAAI Conference on Artificial Intelligence (AAAI) 2021, [paper].