Greg Hamerly

Associate Professor
Computer Science Department
School of Engineering & Computer Science (Rogers building)
105 Baylor Ave.
Baylor University
Waco, TX 76798-7356, USA
Phone: +1-254-710-6846
Fax: +1-254-710-3889
Email: hamerly at cs dot baylor dot edu

I am an associate professor of computer science at Baylor University in Waco, Texas. I'm also the director of graduate studies for computer science. On this webpage you can find more information about my research, teaching, schedule, publications, funding, and other things.

Please note: my family name has one "m". It's "Hamerly", not "Hammerly" (which is the name of different roads in Houston, TX and Dallas, TX, among others), or "Hamerley", "Hammerley", "Hamberly", etc. And my family name is pronounced "hammer-lee".

Research

My research is in machine learning, a sub-field of artificial intelligence. I am interested in unsupervised learning methods (clustering, primarily), and improving algorithms like k-means to be faster, give better-quality results, or to act more intelligently (such as by finding the number of clusters). I have applied machine learning to the task of optimizing computer program simulation, in the SimPoint project. I have done work with mixtures of naive Bayes models in unsupervised learning, for the detection of failures in hard drives. I'm very interested in the automation of learning, especially in the applications knowledge discovery in science, such as bioinformatics and astrophysics.

Here is my NSF-funded research in developing a novel curriculum in computational thinking.

Here is my curriculum vitae, here are my publications, and here is my Ph.D. thesis.

Teaching

Weekly schedule (Fall 2012)

Here is my usual weekly schedule for this semester. If you can't stop by my office during office hours, please email me or just stop by at another time; I am usually glad to meet with students at other times.

Here is my Google calendar that shows when I have scheduled events.

Publications

1. Jonathan Drake, Greg Hamerly. Accelerated k-means with adaptive distance bounds. In OPT2012: the 5th NIPS Workshop on Optimization for Machine Learning, December, 2012. [pdf]
2. William A. Booth, Greg Hamerly, David Sturgill, Ivy Hamerly, Todd Buras. Computational Thinking: Building a Model Curriculum In ACET Journal of Computer Education and Research, 2012. [pdf]
3. Greg Hamerly, Erez Perelman, Timothy Sherwood, Brad Calder, Representative Sampling Using SimPoint. Chapter 10 in the book Processor and System-on-Chip Simulation, edited by Rainer Leupers and Olivier Temam; published by Springer, 2010.
4. Greg Hamerly, Greg Speegle, Efficient Model Selection for Large-Scale Nearest-Neighbor Data Mining In proceedings of the 2010 British National Conference on Databases (BNCOD 2010), June 2010. [pdf]
5. Greg Hamerly, Making k-means even faster In proceedings of the 2010 SIAM international conference on data mining (SDM 2010), April 2010. [pdf]
6. Bing Yin, Greg Hamerly, Hierarchical Stability-Based Model Selection For Clustering Algorithms In proceedings of the International Conference on Machine Learning and Applications, December 2009.
7. Joshua Johnston, Greg Hamerly, Improving SimPoint accuracy for small simulation budgets with EDCM clustering In proceedings of the Second workshop on Statistical and Machine learning approaches to ARchitectures and compilaTion (SMART '08), January 2008. [pdf]
8. Erez Perelman, Jeremy Lau, Harish Patil, Aamer Jaleel, Greg Hamerly, Brad Calder, Cross Binary Simulation Points In Proceedings of the International Symposium on Performance Analysis of Systems and Software (ISPASS-2007) March 2007. [pdf]
9. Yu Feng, Greg Hamerly, PG-means: learning the number of clusters in data. In proceedings of the twentieth annual conference on neural information processing systems (NIPS), December 2006. [ps, pdf]
10. Greg Hamerly, Erez Perelman, Jeremy Lau, Timothy Sherwood, Brad Calder, Using Machine Learning to Guide Architecture Simulation. Journal of Machine Learning Research, Volume 7, Pages 343-378, 2006. [abstract, pdf]
11. Greg Hamerly, Erez Perelman, Brad Calder, Comparing Multinomial and K-means clustering for SimPoint. In the 2006 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS-2006), 2006. [abstract, pdf]
12. Greg Hamerly, Erez Perelman, Jeremy Lau, Brad Calder, SimPoint 3.0: Faster and more flexible program analysis. Journal on Instruction-Level Parallelism (JILP), September, 2005. [pdf]
13. Greg Hamerly, Erez Perelman, Jeremy Lau, Brad Calder, SimPoint 3.0: Faster and more flexible program analysis. Workshop on Modeling, Benchmarking and Simulation (MoBS), June 2005. [abstract, pdf]
14. Brad Calder, Timothy Sherwood, Greg Hamerly, Erez Perelman, SimPoint: Picking Representative Samples to Guide Simulation. Chapter 7 in the book Performance Evaluation and Benchmarking, edited by Lizy Kurian John and Lieven Eeckhout; published by CRC Press, 2005.
15. Jeremy Lau, Erez Perelman, Greg Hamerly, Timothy Sherwood, Brad Calder, Motivation for variable length intervals and hierarchical phase behavior. In the 2005 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS-2005), March 2005. [abstract, pdf]
16. Jeremy Lau, Jack Sampson, Erez Perelman, Greg Hamerly, Brad Calder, The strong correlation between code signatures and performance. In the 2005 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS-2005), March 2005. [abstract, pdf]
17. John Seng, Greg Hamerly, Exploring perceptron-based register value prediction. In the second value-prediction and value-based optimization workshop, October 2004. [pdf]
18. Greg Hamerly, Erez Perelman, Brad Calder, How to use SimPoint to pick simulation points. In ACM SIGMETRICS Performance Evaluation Review, Volume 31(4), March 2004. [abstract, pdf]
19. Tim Sherwood, Erez Perelman, Greg Hamerly, Suleyman Sair, Brad Calder, Discovering and Exploiting Program Phases. In IEEE Micro: Micro's top picks from computer architecture conferences, November-December 2003 (Vol. 23, No. 6) [pdf].
20. Greg Hamerly, Charles Elkan, Learning the k in k-means. In proceedings of the seventeenth annual conference on neural information processing systems (NIPS), pages 281-288, December 2003. [ps, pdf] (Older UCSD technical report CS2002-0716 [ps])
21. Erez Perelman, Greg Hamerly, Brad Calder, Picking Statistically Valid and Early Simulation Points. In proceedings of the international conference on parallel architectures and compilation techniques (PACT), September 2003. [abstract, pdf]
22. Erez Perelman, Greg Hamerly, Michael Van Biesbrouck, Tim Sherwood, Brad Calder, Using SimPoint for Accurate and Efficient Simulation. In proceedings of the international conference on measurement and modeling of computer systems (SIGMETRICS), June 2003. [abstract, pdf]
23. Tim Sherwood, Erez Perelman, Greg Hamerly, Brad Calder, Automatically characterizing large scale program behavior. In proceedings of the tenth international conference on architectural support for programming languages and operating systems (ASPLOS), October 2002. [abstract, pdf]
24. Greg Hamerly, Charles Elkan, Alternatives to the k-means algorithm that find better clusterings. In proceedings of the ACM conference on information and knowledge management (CIKM), pages 600-607, November 2002. [ps] (Older UCSD technical report CS2002-0702 [ps])
25. Greg Hamerly, Charles Elkan, Bayesian approaches to failure prediction for disk drives. In proceedings of the eighteenth international conference on machine learning (ICML), June 2001. [ps]

Here are links to my coauthors and collaborators: Brad Calder, Charles Elkan, Yu Feng, Aamer Jaleel, Jeremy Lau, Harish Patil, Erez Perelman, Jack Sampson, Suleyman Sair, John Seng, Tim Sherwood, Michael Van Biesbrouck.

An upper bound on my Erdös number is 4. One such path is me → Charles Elkan → Russell Greiner → Michael S. O. Molloy → Paul Erdös. Another such path is me → Tim Sherwood → Ömer Eğecioğlu → Charles Ryavec → Paul Erdös.

Thesis

Defended on June 26, 2003. Learning structure and concepts in data through data clustering. [ps, 1.8MB] [pdf, 3.6MB]
Thanks to Tom Stepleton at Sony Japan for catching a typo in one of my equations.

Funding

My research and teaching work has been generously supported by the following:

• National Science Foundation
• Intel Corporation
• Baylor University Young Investigator's Development Program

Current and former students

• Jonathan Drake, M.S. 2013, Undergraduate Scholars Thesis 2011
• Tak-Chien Chiam, M.S. 2012, currently at Amazon
• Hao Guo, M.S. 2012
• George Montanez, M.S. 2011, working on a Ph.D. at Carnegie Mellon
• Winston Ewert, M.S. 2011, working on a Ph.D. at Baylor
• Lei Meng, M.S. 2011, working on a Ph.D. at Notre Dame
• Bing Yin, M.S. 2009, currently at Amazon
• Josh Johnston, M.S. 2007, working on a Ph.D. at Texas A&M
• Yu Feng, M.S. 2006, currently at Microsoft

Current/former Affiliations

• I am a member of ACM
• DTAI group at the KU Leuven computer science department
• AI lab at the UCSD computer science department
• Computer science department at CalPoly, San Luis Obispo

Programming contests

• I went to the ACM ICPC world finals as a contestant for UC San Diego in 2000 and 2001.
• I have been a coach for students at UCSD and Baylor.
• I regularly teach Competitive Learning at Baylor, a course on how to do algorithmic problem solving. I developed this course with David Sturgill, and together we have authored more than 300 problems for the course.
• I participated in putting together the following contests:
  • The 2009 ICPC World Finals Playoff.
  • The 2010, 2011, and 2012 practice problems at the ICPC World Finals.
  • The 2012 North American Qualifier Competition.
• I have been a site director for the ICPC South Central USA regional competition in 2010, 2011, and 2012.
• I have been a member of the ICPC Live Analytics team at the 2011 and 2012 world finals.

Links to other things I've done

• Some sites I frequent are in my bookmarks.
• Here is a list of software I have written and released.

Copyright © 2004 Greg Hamerly
Computer Science Department
Baylor University

valid html and css