Zilong Tan

alt text 

Department of Computer Science
Duke University
LSRC Building D101
308 Research Dr.
Durham, NC 27708
Phone: +1 919 599 3120
E-mail: ztan@cs.duke.edu

About me

I am a postdoctoral associate in the machine learning department at CMU. I work on scalable algorithms for machine learning with applications in computational biology. I obtained a PhD degree in Computer Science from Duke University.

Research

My research projects include

  • Approximate inference using tensor decomposition and randomized algorithms

  • Inference and variable selection in reproducing kernel Hilbert space

  • Sufficient dimension reduction

Publications

  • Learning Compressed Model Representations
    Zilong Tan and Sayan Mukherjee

  • Subspace-Induced Gaussian Processes
    Zilong Tan and Sayan Mukherjee
    [arXiv, Code]
    Submitted

  • Scalable Algorithms for Learning High-Dimensional Linear Mixed Models
    Zilong Tan, Kimberly Roche, Xiang Zhou, and Sayan Mukherjee
    UAI 2018
    [arXiv, Code]

  • Partitioned Tensor Factorizations for Learning Mixed Membership Models
    Zilong Tan and Sayan Mukherjee
    ICML 2017
    [Paper, Slides, Code]

  • Efficient Learning of Graded Membership Models
    Zilong Tan and Sayan Mukherjee
    [arXiv]

  • Tempo: Robust and Self-tuning Resource Management in Multi-tenant Parallel Databases
    Zilong Tan and Shivnath Babu
    VLDB 2016
    [Paper, Code]

Software

  • SIGP: Matlab implementation of subspace-induced Gaussian processes

  • arLMM: fast linear mixed models, useful for GWAS

  • PTPQP: partitioned tensor factorization

  • PALD: multiobjective Pareto optimization

  • fast-hash: a fast hash function “learned” using genetic programming