Yao Xie

yao photo

Yao Xie
Assistant Professor (Official webpage)
Harold R. and Mary Anne Nash Early Career Professor
H. Milton Stewart School of Industrial and Systems Engineering
Georgia Institute of Technology

Georgia Tech Statistics Center
Georgia Tech Center for Machine Learning
Georgia Tech IPaT (Institute for People and Technology)

Phone: (404) 385-1687
Office: Groseclose Building 339
765 Ferst Dr NW, Atlanta, GA, 30332-0205
Email: Email address

Research Interests

Signal processing, statistics, optimization, and machine learning. I am particularly interested in sequential analysis and change-point detection, and their broad applications in sensor networks, social networks, imaging and communications.

Currently Teaching

ISyE 2028, Basic Statistical Methods. Fall 2018. Syllabus, Tentative schedule.


August 2018, two papers accepted for spotlight presentation at NIPS 2018: "Robust hypothesis testing using Wasserstein uncertainty sets", with R. Gao, L. Xie, and H. Xu, and "Learning temporal point processes via reinforcement learning", with S. Li, S. Xiao, S. Zhu, N. Du, and L. Song.

August 2018, Shuang Li is a Finalist in the 2018 INFORMS QSR Best Student Paper Competition. Congratulations Shuang!

August 2018, receive new grants from National Science Foundation (NSF) and the Atlanta Police Foundation (APF). Thank you NSF, and APF!

July 2018, give a seminar at Kyoto University, and talk at Discrete Optimization and Machine Learning Workshop in Tokyo, Japan.

June 2018, invited talk at Mostly OM Workshop at Tsinghua University, Beijing, China; invited talk at SPA 2018 (40th Conference on Stochastic Processes and their Applications) in Gothenburg, Sweden; and invited talk at IWAP 2018 (9th International workshop on Applied Probability) in Budapest, Hungary.

May 2018, paper ``Maximum entropy low-rank matrix recovery'' with S. Mak accepted by IEEE Journal of Selected Topics in Signal Processing. Conference version of the paper was accepted by ISIT 2018.

April 2018, Yang Cao has successfully defended his Ph.D. thesis. Congratulations Dr. Cao!

April 2018, talk at IMA Program, Workshop on Forecasting from Complexity, at University of Minnesota, Twin Cities.

March 13 2018, seminar at the University of Washington, Department of Statistics.

March 2018, talk at the Smart Cities Connect Conference and Expo about our Smart 50 Award project.

March 2018, talk at CISS workshop, Princeton University.

March 2018, give a seminar at Stanford University, ISL/IT seminar: "Robust sequential change-point detection".

Feb. 2018, submitted a new paper "Matrix completion with deterministic pattern - a geometric perspective". Joint work with A. Shapiro and R. Zhang.

Feb. 2018, talk at ITA 2018.

Feb. 2018, attend NSF workshop on "Real-time learning and decision making in dynamical systems."

Feb. 2018, paper accepted by EURASIP Journal on Advances in Signal Processing. "On the effect of model mismatch for sequential Info-Greedy Sensing". Joint work with R. Song and S. Pokutta.

Jan. 2018, our crime data analytics project won the "2018 Smart 50 Award" at the Smart Cities Connect Conference & Expo. The project is funded by the Atlanta Police Foundation and a collaboration with the Atlanta Police Department. Congratulations Shixiang and Shuang, who did most of the work!

Jan. 2018, our paper "Sequential change-point detection via online convex optimization" was accepted by Entropy, Special Issue on "Information Theory in Machine Learning and Data Science". Joint work with Y. Cao, L. Xie and H. Xu. A preliminary conference version was accepted by AISTATS 2018: "On near optimality of one-sample update for sequential joint detection and estimation".

Dec. 2017, our paper appeared in Electronic Journal of Statistics: "Change detection via affine and quadratic detectors". Joint work with V. Guigues, A. Juditsky, A. Nemirovski, and Y. Cao.

Dec. 2017, present our paper at CAMSAP conference: "Sequential detection of low-rank changes using extreme eigenvalues" with L. Xie.

Oct. 2017. Junzhuo Chen was selected as one of the finalists for the 2017 INFORMS QSR best student paper competition for our paper ": An efficient score statistic for spatial-temporal surveillance". Congratulations Junzhuo!

Oct. 2017. Seminar at School of Mathematics, Georgia Tech.

Oct. 2017. Talk at Georgia Statistics Day at Emory University.

Oct. 2017. Invited talk at Allerton conference.

Sept. 2017. Demonstrate our system for crime incidents correlation detection (our poster and video demo) at the "Experience the Smart City" event held by the City of Atlanta.

Sept. 2017. Present the Algorithm for Threast Detection Workshop.

Aug. 2017. Tutorial at 2017 Summer School on "Statistical Learning for Big Data", Shenzhen, China.

Aug. 2017. Present at DDDAS Workshop, Cambridge, MA.

July 2017. Post new papers: "Uncertainty quantification and design for noisy matrix completion - a unified framework" with S. Mak, "On near optimality of one-sample update for sequential joint detection and estimation" with Y. Cao, L. Xie and H. Xu, ": An efficient score statistic for spatial-temporal surveillance" with J. Chen and S.-H. Kim, and "Sequential detection of low-rank changes using extreme eigenvalues" with L. Xie.

June 2017. Give talks at IWSM and ISIT.

June 2017. Offer a session "How computers detect human faces" for Georgia Tech, ISyE Misson Poissble program, an summer outreach program for high school students.

March 2017. Guest speaker at Atlanta Science Festival.

March 2017. Seminar at Georgia Tech Center for Machine Learning.

March 2017. I am co-organizing a special session "Sequential methods for high-dimensional structured signals" at ICASSP 2017 (co-organizer: V. V. Veeravalli).

Feb. 2017. Present at ITA at University of San Diego.

Feb. 2017. Seminar at University of Michigan, Ann Arbor, MIDAS, and attended the Women in Data Science Workshop.

Jan. 2017, papers submitted "Change detection via affine and quadratic detectors" (arXived) with V. Guigues, A. Juditsky, A. Nemirovski, and Y. Cao, and "Robust sequential change-point detection by convex optimization" with Y. Cao.

Dec. 2016, our paper "Detecting weak changes in dynamic events over networks" (with Shuang Li, Mehrdad Farajtabar, and Le Song) accepted by IEEE Transactions on Signal and Information Processing over Networks.

Nov. 2016, received Georgia Tech Serve-Learn-Sustain Fellowship for Smart Cities and Connected Communities.

Oct. 2016, receive the Georgia Tech IMat Seed Grant Award. Our proposal is entitled "Dynamically responsive scanning diffraction for high-throughput analysis of phase assemblage in functional complex oxides", and the my role is to process the high-dimensional sequential data. This project is collaborating with J. Kacher and M. Losego at the School of Material Science, Georgia Tech.

Oct. 2016, present at the Georgia Statistics Day Conference.

Sept. 2016, present two papers at the Allerton conference: "Seismic event picking via sequential change-point detection" (with S. Li, Y. Cao, C. Leamon, L. Shi, and W. Song) and "Sequential rank change detection" (with L. Seversky).

Sept. 2016, attend SAMSI workshop on Distributed and Parallel Data Analysis.

Sept. 2016, present at South Big Data Hub workshop, Applications of Analytics and Machine Learning in Energy.

July 2016, one paper accepted by Asilomar conference: "Dynamic change-point detection using correlation networks" (with S. Cao and Y. Chen).

June 2016, I gave an invited talk at ICSA 2016, and an invited talk at COMPSAC 2016.

April 2016, Shuang Li received the 2016 Alice and John Jarvis Ph.D. Student Research Award Honorable Mention. Congratulations Shuang!

April 2016, awarded Visiting Faculty Research Program (VFRP) to work in Air Force Research Lab, Information Directorate in summer.

April 2016, new arXiv paper "Detecting weak changes in dynamic events over networks" (with Shuang Li, Mehrdad Farajtabar, and Le Song); paper "Multi-sensor slope change detection", (with Yang Cao, and Nagi Gebraeel), was accepted by Annals of Operations Research.

March 2016, Christina Leamon's abstract "Change detection of tectonic plate movement in Santorini, Greece" was accepted for an oral presentation at the 11th Annual Undergraduate Research Symposium at Georgia Tech.

March 2016, gave an invited talk at the CISS conference at Princeton.

February 2016, gave an invited talk at the 2016 ITA workshop held at UCSD.

February 2016, gave a seminar at UIUC, ECE department.

Oct. 2015, Yang Cao received the GA Statistics Day Conference 2015 Poster Presentation Award for his poster on Poisson matrix completion. Congratulations Yang!

Nov. 2015, presented at Asilomar conference and be Session Chair. I will also visit Yahoo Research Lab.

Oct. 2015, was elected to be a member of IEEE Machine Learning for Signal Processing Technical Committee (MLSP TC) for a term 2016-2019, and gave a invited talk at Allerton conference. Our paper "Poisson matrix recovery and completion" was accepted by IEEE Transactions on Signal Processing.

Aug. 2015, our research on seismic imaging is featured in a special report on IEEE Signal Processing Magazine. I was included in a Georgia Tech Engineering School News article.

July 2015, Yang Cao was selected to receive a Robert Goodell Brown Fellowship for ISyE in recognition of his strong academic and research achievements. Congratulations Yang!

May-Sept. 2015, submitted several papers: "M-statistic for kernel change-point detection" with S. Li, H. Dai and L. Song (conference version accepted at NIPS 2015), "Multi-sensor slope change detection" with Y. Cao (conference version accepted at Allerton 2015), "Online supervised subspace tracking", with R. Song, H. Dai, Q. Li and L. Song (conference version accepted at Asilomar 2015), "Categorical matrix completion" with Y. Cao (conference version accepted at CAMSAP 2015), and "Sketching for sequential change-point detection" with M. Wang and A. Thompson (conference version accepted at GlobalSIP 2015). "Seismic event picking via sequential change-point detection", with Y. Cao, S. Li, C. Leamon, L. Shi, and W.-Z. Song.

July 2015, our proposal got funded: NSF CMMI-1538746, Collaborative Research: Combining statistical process control and optimization via simulation for robust sensor network design in the presence of sensor measurement error, with PI: Seong-Hee Kim (Georgia Tech) and Co-PI: Aral Mustafa (Georgia Tech).

Aug. 2015, gave a statistics seminar at UGA. June 2015, presented at ISIT. June 14-19, presented at IWSM. June 22-24, gave seminars at Tsinghua University, the Chinese Academy of Sciences, and the Chinese University of Hong Kong (CUHK).

Feb. 2015, attend Algorithms, Radio, and Coding Workshop in honor of Prof. Robert Calderbank's 60th birthday, and ITA 2015. As a tradition of ITA, here's trailer of my talk.