Qi Lei (雷琦)

Qi Lei UT Austin 

Email: leiqi at ices.utexas.edu

Website: http://users.ices.utexas.edu/~leiqi/

I am a Ph.D candidate at Oden Institute for Computational Engineering and Sciences, University of Texas at Austin, working with Professor Alexandros G. Dimakis and Inderjit S. Dhillon. I am also a member of Center for Big Data Analytics and Wireless Networking & Communications Group. I'm visiting IAS for the Theoretical Machine Learning program from September to December this year. Prior to that, I was a Research Fellow at the Simons Institute for the Theory of Computing at UC Berkeley for the program on Foundations of Deep Learing.

My research interests are machine learning, deep learning and optimization. Specifically, I'm interested in developing provably efficient and robust algorithms for some fundamental machine learning problems.

I'm currently looking for postdoc positions starting fall 2020. (Resume, Github, Google Scholar)

News and Announcement

11/2019 I am invited to give a talk on Deep Generative models and Inverse Problems at the mini-symposium “Machine Learning for Solving Partial Differential Equations and Inverse Problems” of the 2nd Annual Meeting of the SIAM Texas-Louisiana Section in Dallas on Nov 2nd, 2019

10/2019 I am going to attend the Rising Stars 2019 at UIUC (An Academic Career Workshop for Women in EECS) from Oct 29th to Nov 1st, 2019

09/2019 I am participating the “Special Year on Optimization, Statistics, and Theoretical Machine Learning” as a short-term visitor at the Institute of Advanced Study from September to December, 2019

09/2019 Two papers accepted to NeurIPS 2019

05/2019 I received Simons-Berkeley fellowship in Foundations of Deep Learning, 2019.

04/2019 Our sysML paper on paraphrasing attacks “Discrete Adversarial Attacks and Submodular Optimization with Applications to Text Classification” was covered by Nature Story, VectureBeat, and TechTalks (code, slides)

Education

Unversity of Texas at Austin, Austin, TX

Ph.D student in Institute for Computational Engineering and Sciences       August 2014 - Present

Zhejiang University, Zhejiang, China

B.S. in Mathematics        August 2010 - May 2014

Experience

Institute of Advanced Study: Theoretical Machine Learning program, Princeton, NJ

Visiting Graduate Student        September 2019 - December 2019

Simons Institute: Foundation of Deep Learning Program, Berkeley, CA

Research Fellow       May 2019 - August 2019

Facebook Visual Understanding Team, Menlo Park, CA

Software Engineering Intern        June 2018 - September 2018

Amazon/A9 Product Search Lab, Palo Alto, CA

Software Development Intern, Search Technologies       May 2017 - August 2017

Amazon Web Services (AWS) Deep Learning Team, Palo Alto, CA

Applied Scientist Intern        January 2017 - April 2017

IBM Thomas J. Watson Research Center, Yorktown Heights, NY

Research Summer Intern        May 2016 - October 2016

UCLA Biomath Department, Los Angeles, CA

Visiting Student        July 2013 - September 2013

Preprint

17. Qi Lei, Jason D. Lee, Alexandros G. Dimakis, Constantinos Daskalakis. “SGD Learns One-Layer Networks in WGANs”, arxiv preprint

Publications

16. Qi Lei, Jiacheng Zhuo, Constantine Caramanis, Inderjit S Dhillon, Alexandros G Dimakis. “Primal-Dual Block Frank-Wolfe”, To appear in NeurIPS 2019

15. Qi Lei, Ajil Jalal, Inderjit S. Dhillon, Alexandros G. Dimakis. “Inverting Deep Generative models, One layer at a time”, To appear in NeurIPS 2019

14. Qi Lei, Jinfeng Yi, Roman Vaculin, Lingfei Wu, Inderjit Dhillon. “Similarity Preserving Representation Learning for Time Series Analysis”, The 28th International Joint Conference on Artificial Intelligence (IJCAI), 2019. (code)

13. Qi Lei, Lingfei Wu, Pin-Yu Chen, Alexandros G. Dimakis, Inderjit S. Dhillon, Michael Witbrock. “Discrete Adversarial Attacks and Submodular Optimization with Applications to Text Classification”, Systems and Machine Learning (sysML). 2019 (code, slides)

12. Jinfeng Yi, Qi Lei, Wesley Gifford, Ji Liu. “Negative-Unlabeled Tensor Factorization for Location Category Inference from Inaccurate Mobility Data”, SIAM International Conference on Data Mining (SDM), 2019 (code)

11. Zhewei Yao, Amir Gholami, Qi Lei, Kurt Keutzer, Michael W. Mahoney. “Hessian-based Analysis of Large Batch Training and Robustness to Adversaries”, Neural Information Processing Systems (NIPS), 2018

10. Jiong Zhang, Qi Lei, Inderjit Dhillon, “Stabilizing Gradients for Deep Neural Networks via Efficient SVD Parameterization”, International Conference of Machine Learning (ICML), July. 2018

9. Lingfei Wu, Ian En-Hsu Yen, Jinfeng Yi, Fangli Xu, Qi Lei and Michael Witbrock, “Random Warping Series: A Random Features Method for Time-Series Embedding”, Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics (AISTATS), 2018

8. Hsiang-fu Yu, Cho-Jui Hsieh, Qi Lei, Inderjit Dhillon, “A Greedy Approach for Budgeted Maximum Inner Product Search”, Proc. of Neural Information Processing Systems (NIPS), 2017

7. Qi Lei, Enxu Yan, Chao-yuan Wu, Pradeep Ravikumar, Inderjit Dhillon, “Doubly Greedy Primal-Dual Coordinate Methods for Sparse Empirical Risk Minimization”, Proc. of International Conference of Machine Learning (ICML), 2017 (code)

6. Rashish Tandon, Qi Lei, Alexandros G. Dimakis, Nikos Karampatziakis, “Gradient Coding: Avoiding Stragglers in Distributed Learning”, Proc. of International Conference of Machine Learning (ICML), 2017 (code)

5. Qi Lei, Kai Zhong, Inderjit. Dhillon, “Coordinate-wise Power Method”, Proc. of Neural Information Processing Systems (NIPS), Dec. 2016 (code,poster)

4. Arnaud Vandaele, Nicolas Gillis, Qi Lei, Kai Zhong, Inderjit Dhillon, “Efficient and Non-Convex Coordinate Descent Methods for Symmetric Nonnegative Matrix Factorization”, IEEE Transactions on Signal Processing 64.21 (2016): 5571-5584 (code)

3. Maria R. D'Orsogna, Qi Lei, Tom Chou, “First assembly times and equilibration in stochastic coagulation-fragmentation”, The Journal of Chemical Physics, 2015: 143.1, 014112

2. Jiazhou Chen, Qi Lei, Yongwei Miao, Qunsheng Peng, “Vectorization of Line Drawing Image based on Junction Analysis”, Science China Information Sciences, 2014:1-14 (code)

1. Jiazhou Chen, Qi Lei, Fan Zhong, Qunsheng Peng, “Interactive Tensor Field Design Based on Line Singularities”, Proceedings of the 13th International CAD/Graphics, 2013 (code)

Service

conf review: AAAI’20 (PC), NIPS’19, ACML’19 (PC), ICLR’19, ICML’19, AISTATS’19, NIPS’18, ICML’18, AISTATS’18, NIPS’16, and more

journal review: MOR’19, TNNLS’19, TKDE’19, ISIT’18, TIIS’17, IT’17, ISIT’17, IT’16, and more

Teaching

Scalable Machine Learning, Teaching Assistant, Fall 2019

Mathematical Methods in Applied Engineering and Sciences, Instructer Intern, Spring 2016

Patent

“Method and System for General and Efficient Time Series Representation Learning via Dynamic Time Warping”
with J. Yi, R. Vaculin, W. Sun

“Real-Time Cold Start Recommendation and Rationale within a Dialog System”
with J. Yi, R. Vaculin, M. Pietro