Ji Liu


Ji Liu

Ji Liu

Assistant Professor

  • Department of Computer Sciences
  • Department of Electrical and Computer Engineering
  • University of Rochester
  • Rochester, NY 14627
  • Office: CSB 614
  • Email: jliu AT cs.rochester.edu




News

  • I just received the 2017 IBM faculty award. Thanks all group members for their efforts and the generosity of IBM.
  • Our Asynchronous machine learning solver has been released over here.
  • http://junehuang.com/software.html
  • Two papers get accepted by ICML 2017 in our group. Congratulations to Haichuan, Shupeng, and other authors.
  • Our Machine Teaching paper gets accepted by IJCAI 2017. Jerry and Manuel, thank you guys for expanding this new area.
  • Our finite-sum composition optimization paper gets accepted by AISTATS 2017. Congratulations to Xiangru Lian and thank Professor Mengdi Wang.
  • Our paper ``Prognostics of Surgical Site Infections using Dynamic Health Data'' is accepted by Journal of Biomedical Informatics. Congratulations to all authors, especially the leading author Chuyang Ke. He is an undergraduate working with me.
  • Three papers get accepted by NIPS 2016 in our group. Congratulations to all authors, especially Xiangru Lian. He has two papers accepted as the leading authors this year.
  • Our paper "Online Feature Selection: A Limited-Memory Substitution Algorithm and its Asynchronous Parallel Variation" gets accepted by KDD 2016. Congratulations to Haichuan Yang and other authors. This is Haichuan's second top conference paper as the first author in this semester.
  • Our paper "Optimal Denoising Matrix in Dantzig Selector" gets accepted by UAI 2016. Congratulations to all authors.
  • Our paper "The Teaching Dimension of Linear Learners" gets accepted by ICML 2016. Congratulations to all authors. For the first time, this paper gives the lower bound of required teaching items to teach a linear learner.
  • Two papers get accepted on IJCAI 2016. Congratulations to all authors.
  • Present our work "Asynchronous parallel stochastic gradient work for nonconvex optimization" at Princeton University on Informs Optimization 2016.
  • Our paper "On Benefits of Selection Diversity via Bilevel Exclusive Sparsity" gets accepted in CVPR 2016. This is the first paper to theoretically justify the benefits of enforcing diversity in selection. Congratulations to all authors.
  • Three papers get accepted in AAAI 2016. Congratulations to all authors.
  • Our paper "Asynchronous parallel stochastic gradient for nonconvex optimization" was accepted by NIPS 2015 as spotlight paper. Congratulations to all authors.
  • Our paper "Finite-Sample Analysis of GTD Algorithms" receives the BEST STUDENT PAPER award in UAI 2015.
  • Congratulations to Iftekhar for publishing his paper "Unsupervised Extraction of Human-Interpretable Nonverbal Behavioral Cues in a Public Speaking Scenario" in ACMMM with me and Ehsan. This paper is based on his course project in my class CSC 576.
  • Our paper "Finite-Sample Analysis of GTD Algorithms" is accepted in UAI 2015 as PLENARY presentation.
  • Be serving IJCAI 2015 as a Senior PC.
  • Our paper ``Exclusive Feature Learning on Arbitrary Structures'' is accepted by NIPS, 2014.
  • Our paper ``An Asynchronous Parallel Stochastic Gradient Descent'' is accepted by JMLR, 2014.
  • Our paper (with Y. Cong and J. Luo) is accepted by TCSVT, 2014.
  • Our paper (with M. Collins, J. Xu, L. Mukherjee, and V. Singh) is accepted in ECCV 2014.
  • SIAM conference on optimization: Present our work ``Asynchronous Parallel Stochastic Coordinate Descent'' in San Diego, May, 2014
  • Two first author papers are accepted in ICML 2014.

Education

  • Ph.D., Computer Science, University of Wisconsin-Madison (advisor Stephen J. Wright)
  • M.S., Computer Science, Arizona State University (advisors Jieping Ye and Peter Wonka)
  • B.S., Automation (major) and Business Administration (minor), University of Science and Technology of China (USTC)

Research Interests

  • Optimization and Data Analytics (parallel methods for big data applications, large scale problems, first order algorithms, stochastic algorithms, bioinformatics)
  • Machine Learning (compressed sensing theory, tensor completion / decomposition, multi-tasks learning, reinforcement learning)
  • Applications (healthcare, bioinformatics, spectral clustering, abnormal event detection, online metric learning, dictionary learning, image segmentation, image in-painting)

Honors

  • IBM faculty award, 2017
  • Best student paper award in UAI, 2015
  • Honorable mention of the best research paper in KDD, 2010
  • ICML Scholarship, 2013
  • NIPS Travel Grant, 2010
  • PAMI TC Travel Grant, 2009
  • Grade 3, National Graduate Mathematical Contest in Modeling, 2007
  • First Prize, Mathematics Contest in Modeling of Northeast China, 2007 (7/2000+)
  • Excellent Student Scholarship, Chinese Academy of Sciences, 2007
  • Graduate Fellowship with Honor, Chinese Academy of Sciences, 2006
  • Outstanding Student Scholarship, USTC, 2003, 2004

Services

  • PC in NIPS 2013, 2014, 2015, 2016
  • PC in ICLR 2016
  • PC in SDM 2015, 2016
  • SPC in IJCAI 2015
  • Reviewer for Annals of Statistics (AoS)
  • Reviewer for SIAM on Optimization (SIOPT)
  • Reviewer for Journal of Machine Learning Research (JMLR)
  • Reviewer for Mathematical Programming (MP)
  • Reviewer for IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI)
  • Reviewer for IEEE Transactions on Information Theory (TIT)
  • Reviewer for Journal of the ACM (JACM)
  • Reviewer for Numerical Algorithms (NUMA)
  • Reviewer for IEEE Transactions on Signal Processing (T-SP)
  • Reviewer for IEEE Transactions on Neural Networks and Learning Systems (T-NNLS)
  • Reviewer for IEEE Transactions on Knowledge and Data Engineering (T-KDE)
  • Reviewer for Computer Vision and Image Understanding, Elsevier (CVIU)
  • Reviewer for Neurocomputing
  • Reviewer for Digital Signal Processing, Elsevier (DSP)
  • Reviewer for International Journal of Engineering, Science and Technology (IJEST)

Invited Talks

  • ISMP talk, "Forward-backward greedy algorithms for convex smooth optimization under a cardinality constraint", Pittsburgh, July, 2015
  • Yahoo talk, ``Feature Selection via Sparse Learning'', Sep., 2015
  • NEC talk, ``Asynchronous parallel optimization'', Sep., 2015
  • Vision And Learning SEminar talk, "Feature Selection via Sparse Learning", Online. July, 2015
  • Xerox talk, "Feature Selection", Rochester. Dec. 2014
  • Symposium on Coordinate Descent Methods @ SIAM Conference on Optimization, San Diego, May, 2014
  • INFORMS Optimization Society, Conference, session "Algorithms for large-scale structured optimization and applications", Houston, Mar. 2014
  • Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, China, Dec., 2011