Avishek Saha

School of Computing
University of Utah

Contact info:
50 S. Central Campus Drive, Rm.3190,
Salt Lake City, UT 84112-9205

office: Rm.3383 (Merrill Engineering Building)
phone: 801-581-4867
email: firstname at cs dot utah dot edu



Bio | Publications | All Publications | More Acads Stuff


Bio

I am a Computer Science PhD student [2007-till date) at School of Computing, University of Utah. I am lucky to have Suresh Venkatasubramanian as my advisor. In addition, I am fortunate enough to be able to collaborate with Hal Daumé III and Jeff M. Phillips.

I am primarily interested in machine learning and have done some work in the area of transfer learning. Currently, I am working on a distributed model for machine learning. Prior to this, I have dabbled in databases, video compression and computer architecture.

I did my Masters [2005-2007] in Computer Science from Indian Institute of Technology, Kharagpur and my Bachelors [1999-2003] in Electronics and Telecommunications Engineering from Jadavpur University, India. In between, I worked for a couple of years on software development for embedded systems. My masters thesis was on motion estimation algorithms for video compression and my bachelors thesis involved navigational path planning for mobile robots.

I have spent fun summers interning at NEC Labs (Machine Learning group) [June-September, 2011] and at AT&T Shannon Labs (Database group) [June-August, 2008].

I am a recipient of the 2011-2012 University of Utah Graduate Research Fellowship. Thanks to the university for sponsoring my research.


Publications
[J:Journals, C:Conferences, X:arXiv, W:Workshops, P:Patents | Items marked * have their author list in alphabetical order]

    Preprints

  1. *[X] Fast Algorithms for Multiple Kernel Learning using Multiplicative Weight Updates
    by John Moeller, Parasaran Raman, Avishek Saha, Suresh Venkatasubramanian
    [abstract] [paper] [code]

    Postprints

  2. *[C,X] Efficient Protocols for Distributed Classification and Optimization
    by Hal Daumé III, Jeff M. Phillips, Avishek Saha, Suresh Venkatasubramanian
    at Algorithmic Learning Theory (To appear), 2012.
    [abstract] [arXiv] [also see a concurrent but independent work]

  3. *[C,X] Protocols for Learning Classifiers on Distributed Data
    by Hal Daumé III, Jeff M. Phillips, Avishek Saha, Suresh Venkatasubramanian
    at International Conference on Artificial Intelligence and Statistics (AISTATS), 2012.
    [abstract] [paper] [arXiv(full version)]

  4. [C] Active Supervised Domain Adaptation
    by Avishek Saha, Piyush Rai, Hal Daumé III, Suresh Venkatasubramanian, Scott L. DuVall
    at European Conference on Machine Learning (ECML), 2011.
    [abstract] [paper]

  5. [C] Online Learning of Multiple Tasks and Their Relationships
    by Avishek Saha, Piyush Rai, Hal Daumé III, Suresh Venkatasubramanian
    at International Conference on Artificial Intelligence and Statistics (AISTATS), 2011.
    [abstract] [paper]

  6. *[C] Co-regularization Based Semi-supervised Domain Adaptation
    by Hal Daumé III, Abhishek Kumar, Avishek Saha
    at Advances in Neural Information Processing Systems (NIPS), 2010.
    [abstract] [paper] [supplement]

  7. [W] Active Online Multitask Learning
    by Avishek Saha, Piyush Rai, Hal Daumé III, Suresh Venkatasubramanian
    at Budgeted Learning Workshop (in conjunction with ICML 2010), 2010.
    [abstract] [paper]

  8. *[W] Frustratingly Easy Semi-supervised Domain Adaptation
    by Hal Daumé III, Abhishek Kumar, Avishek Saha
    at ACL 2010 Workshop on Domain Adaptation for Natural Language Processing (DANLP), 2010.
    [abstract] [paper]

  9. [W] Domain Adaptation meets Active Learning
    by Piyush Rai, Avishek Saha, Hal Daumé III, Suresh Venkatasubramanian
    at Active Learning for NLP (ALNLP), Workshop at NAACL-HLT 2010, 2010.
    [abstract] [paper]

  10. [W] Co-regularization Based Analysis of Feature Sharing Algorithms
    by Abhishek Kumar, Avishek Saha, Hal Daumé III, Tom Fletcher, Suresh Venkatasubramanian
    in Learning Workshop, 2010, 2010.
    [extended abstract]

  11. *[C] Sequential Dependencies
    by Lukasz Golab, Howard Karloff, Flip Korn, Avishek Saha, Divesh Srivastava
    in 35th International Conference on Very Large Data Bases (VLDB09), 2009.
    [abstract] [paper]

  12. *[P] Processing Data Using Sequential Dependencies
    by Lukasz Golab, Howard Karloff, Flip Korn, Avishek Saha, Divesh Srivastava (Patent Filed 2009)
    [based on the paper Sequential Dependencies (VLDB09)].

  13. *[C] Metric Functional Dependencies
    by Nick Koudas, Avishek Saha, Divesh Srivastava, Suresh Venkatasubramanian
    in 25th International Conference on Data Engineering (ICDE09), 2009.
    [abstract] [paper]

Click here for the complete list.


More Acads Stuff

Reviewer:

I am a regular member of: In addition, I have been a teaching assistant for the courses: Relevant Coursework:


Locations of visitors to this page
last updated: february 2012