Home Research Favorite Hobby Album Contact

Ting Liu

The printable version of my resume: [PDF] [PS]

Computer Science Department
Carnegie Mellon University
5000 Forbes Ave
Pittsburgh, PA 15213
  5530 Fifth Ave, Apt 3B
Pittsburgh, PA, 15232
Email: tingliu AT cs DOT cmu DOT edu
Web page:http://www.cs.cmu.edu/~tingliu


To find a permanent position in research labs, research-oriented companies, or financial corporations.

Research Interests

My research interest lies in machine learning and data mining, with focus on nonparametric statistics, memory-based learning and kernel-based learning. I am currently interested in designing high-performance algorithms that solve fundamental tasks (such as k nearest neighbor and support vector machine) on massive and high-dimensional data sets. I am also interested in solving real-world problems, such as content-based vedio segmentation, image retrieval, biological problems and financial modeling problems.


  • Carnegie Mellon University,Pittsburgh, PA. 2001-Present
    Ph.D. Candidate in Computer Science Department.
    Adviser: Andrew W. Moore

  • Tsinghua University, Beijing, P.R.China, 1997-2001.
    B.E. in Computer Science, June 2001.
  • List of Publications (thesis related)

  • T.Liu, A. W. Moore, A. Gray.
    New Algorithms for Efficient High-Dimensional Nonparametric Classification.
    Accepted to Journal of Machine Learning Research.

  • T.Liu, A. W. Moore, A. Gray.
    New Algorithms for Efficient High-Dimensional Nonparametric Classification.
    To appear in Chapter 4, Nearest-Neighbor Methods in Learning and Vision.

  • J. Adcock, A. Girgensohn, M. Cooper, T. Liu, E. Rieffel, L. Wilcox.
    FXPAL Experimentx for TRECVID 2004.
    In Proceedings of TRECVID 2004, March 1, 2005.[PDF]

  • T.Liu, A. W. Moore, A. Gray, Ke Yang.
    An Investigation of Practical Approximate Nearest Neighbor Algorithms.
    In proceedings of Neural Information Processing Systems(NIPS 2004), Vancouver, BC, Canada, 2004.[PS][PDF][Dataset]

  • T.Liu, Ke Yang, A. W. Moore.
    The IOC algorithm: Efficient Many-Class Non-parametric Classification for High-Dimensional Data.
    In proceedings of the 10th ACM SIGKDD conference(KDD 2004),Seattle,WA,2004.[PS][PDF]

  • A.Gray,P.Komarek,T.Liu,A.W.Moore.
    High-Dimensional Probabilistic Classification For Drug Discovery .
    In proceedings of COMPSTAT 2004, 16th Symposium of IASC,Prague,2004.[PS][PDF]

  • T.Liu, A. W. Moore, A. Gray.
    Efficient Exact k-NN and Nonparametric Classification in High Dimensions.
    In Proceedings of Neural Information Processing Systems, 2003.[PS][PDF]

  • Y. Qi, A. Hauptman, T.Liu.
    Supervised Classification for Video Shot Segmentation.
    In proceedings of IEEE International Conference on Multimedia & Expo, 2003.[PDF]

    List of Publications (something else I did for fun)

  • C. Lin, Z. Shan, T. Liu, Y. Qu, F Ren.
    Modeling and Inference of Extended Interval Temporal Logic for Nondeterministic Intervals.
    In IEEE Transactions On Systems, Man, and Cybernetics, 2005, 35(5):682-696.

  • T.Liu, C. Lin, W. Liu.
    Linear temporal inference of workflow management system based on timed Petri net models.
    In Acta Electronica Sinica, Feb. 2002, 30(2): 245-248.

  • T.liu, C. Lin, W. Liu.
    The inference engine of extended interval temporal logic.
    In Chinese Journal of Computers, 2002, 25(6):637-644.

  • C. Lin, T.Liu, Y. Qu.
    Extended interval temporal logic for undetermined Interval:modeling and linear inference using time Petri nets.
    In Chinese Journal of Computers},2001, 24(12):1299-1309.

    Professional Experiences

  • Summer Intern (June, 2005 -- Oct, 2005): Google Inc.
  • My work in Google mainly focused on near-duplicate image search problem. By extending a previous research project, we developed very efficient and scalable algorithm capable of finding approximates nearest neighbors and clustering more than 1 billion images.
    References: Dr. Henry Rowley, Dr. Chuck Rosenberg

  • Summer Intern (June,2004 -- Sept,2004): Fuji Xerox Palo Alto Laboratory
  • My work in FXPAL mainly focused on building a simple general framework for automatic Video Shot Segmentation, and we submit our result to TRECVID04 evaluation, and ranks 2nd among 35 submissions from world-wide participants.
    Reference:: Dr. Matthew Cooper, Dr. Eleanor Rieffel

  • Technical Consultant (Sept,2004 -- Sept, 2005): Fuji Xerox Palo Alto Laboratory
  • We extend our work of shot segmentation to story based segmentation.
    References: Dr. Matthew Cooper, Dr. Eleanor Rieffel


    M. Cooper, T. Liu, E. Rieffel
    Media Segmentation Combining Similarity Analysis and Classification, Filed 11/12/2004

    Research Experience

  • Image Retrieval (On going)
    We are working on detecting whether an image is a near duplicate or a sub image of a database of images.
    References: Prof. Martial Hebert, Ke Yan.

  • 3-D models for Computer Vision
    One popular approach to recognizing objects in 3D data is to use semi-local shape signatures to find similarly-shaped regions between a scene and objects from a model database, conventional algorithm is slow, we are investigating the application of our new data-structure to further enhance recognition speed.
    References: Prof. Martial Hebert, Dr. Daniel Huber.

  • Efficient Nonparametric Classification in High Dimensions
    We designed new ball tree algorithms to achieve non-approximate acceleration of high dimensional nonparametric operations such as k nearest neighbor classifiers and the prediction phase of Support Vector Machine classifiers.
    References:Prof. Andrew W.Moore, Prof. Alexander Gray.

  • Pfizer Global Research and Development
    Computational chemistry and proteomics for drug design. High-throughput screening (metric learning, classification).
    References: Prof. Andrew W. Moore, Prof. Alexander Gray.

    Course Projects

    Computer Vision. Sep. 2004 -- Dec. 2004
  • Shape Matching and Object Recognition Using Shape Contexts.
    We implement a novel approach to measure similarity between shapes and exploit it for object recognition.
    Reference: Prof. Martial Hebert.
  • Computer Network. Sep. 2002 -- Dec. 2002

  • Geographic Routing with Imperfect Positioning for Wireless Networks.
    We proposed a new routing protocol that works well in ad-hoc wireless networks with highly dynamic topology and imprecise geographic position information.
    Reference: Prof. Srini Seshan.
  • Computer Architecture. Sep. 2001 -- Dec. 2001

  • Evaluating Dynamic Prefetching in a Software DSM (Distributed Shared Memory) system.
    We proposed and evaluated a dynamic prefetching algorithm in JIAJIA (a software DSM system).
    Reference: Prof. Todd C. Mowry.
  • Teaching Experience

  • Teaching Assistant for "15-381: Artificial Intelligence", Spring 2005, Carnegie Mellon University.
  • Teaching Assistant for "15-780/16-731, Anvanced AI concepts / Fundamentals of AI for Robotics", Spring 2004, Carnegie Mellon University.
  • Awards and Hornors

  • Carnegie Mellon University Doctorate Fellowship, 2001 -- Present.
  • Graduated with Honor from Tsinghua University, 2001.
  • Tsinghua-ISS Scholarship for outstanding students, 2000.
  • Beijing outstanding college student of the year, 1999.
  • Tsinghua-Sony Scholarship for outstanding students, 1998.
  • First Prize in National Mathematics Contest (China), 1995.
  • References

    Available upon request.