Statistical Machine Learning and Visualization Lab

 School of Computational Science and Engineering
 College of Computing
 Georgia Institute of Technology, Atlanta GA

Overview

The Statistical Machine Learning and Visualization Lab is a research group focused on machine learning and visualization of high dimensional data. Our research emphasizes statistics and computation, and includes both basic research and applied studies.

We are a part of the school of computational science and engineering within Georgia Tech's College of Computing. Our physical location is Klaus 1305/1308. A related CSE lab that is also focused on machine learning is the FAST lab. The FODAVA initiative carries out related research.

People

Faculty Frequent Collaborators
Guy Lebanon Saurabh Bagchi (Purdue)
  Kevyn Collins-Thompson (Microsoft)
Students Pinar Donmez (CMU)
Krishnakumar Balasubramanian Paul Kidwell (LLNL)
Joshua Dillon Yang Zhao (Google)
Seungyeon Kim  
Yi Mao  
Mingxuan Sun  

Papers
  • Lists of publications are available from the individual members websites. Most papers are listed here.

News and Announcements
  • We have an opening for a postdoc position in machine learning. To apply please email Prof. Lebanon a CV, contact information for 2-3 letter writers, and 3 representative publications and specify the email subject as "postdoc application".
  • Seungyeon Kim is a new MS student who is joining our lab. He will work on sequential visualization of documents.
  • February and March have a bunch of other deadlines: SIGIR (1/22), KDD (2/5), ICML (2/1), ACL (2/15), UAI (3/11).

Hosted Visitors and Related Talks
  • Kevyn Collins-Thompson (Microsoft Research) 12/4/09
  • Yoram Singer (Google Research) 2/12/10
  • Max Welling (UCI) 3/19/10
  • Lillian Lee (Cornell) 4/9/10
  • Joseph S. Verducci (Ohio State University) 4/23/10
  • Robert Schapire (Princeton) 4/30/10
  • Susan Holmes (Stanford) TBD
Unless noted otherwise, seminars are Fridays 2-3pm in KACB 2447.

Active Projects

Computationally efficient parameter estimation in graphical models [1]
Dillon, Lebanon

Non-parametric modeling of concept drift [1,2]
Lebanon, Zhao, Zhao

Generative semi-supervised learning
Balasubramanian, Dillon, Lebanon

Incorporating domain knowledge into statistical modeling [1]
Lebanon, Mao

Machine Learning for Language Processing [1,2,3,4,5,6,7,8]
Collins-Thompson, Dillon, Lebanon, Mao

Analysis of Computer Systems Data [1,2]
Bagchi, Modelo-Howard, Lebanon

Modeling and analysis of preference data [1,2,3]
Kidwell, Lebanon, Mao, Sun

Non-parametric approaches to collaborative filtering
Kidwell, Lebanon, Sun

Unsupervised estimation of supervised risks [1]
Balasubramanian, Donmez, Lebanon

Visualizing text documents [1,2,3]
Balasubramanian, Dillon, Kim, Lebanon, Mao

Visualizing search engines [1]
Collins-Thompson, Lebanon, Sun

Funding

Our lab is funded primarily by Georgia Tech and the US National Science Foundation