Kevin L. Chang
About me
I was a postdoctoral fellow at the Max Planck Institute for Computer Science
in Germany. I recently finished a PhD from the
Computer Science Department
at Yale. Previously, I was an undergraduate math major at
Princeton.
Research
My many research interests include:
- Massive data set algorithms and data mining
- Optimization
- Machine learning
My thesis. (Please read!)
[Postcript version]
[PDF version]
Publications
-
Approximation schemes for degree-restricted MST and Red-Blue
separation problem.
with
Sanjeev Arora.
Algorithmica, 40(3):189-210, 2004.
A preliminary version appeared in
ICALP 2003.
[Postscript of journal version]
-
Inoculation strategies for victims of viruses and sum-of-squares partition problem.
with James Aspnes and Aleksandr Yampolskiy.
Journal of Computer and System Sciences, 72(6):1077-1093, 2006.
A preliminary version appeared in SODA 2005.
[Postscript of journal version]
[Postscript of SODA version]
[PowerPoint slides]
-
The Space Complexity of Pass Efficient Algorithms for Clustering.
with Ravi Kannan.
To appear in SIAM Journal on Computing.
A preliminary version appeared in SODA 2006.
[Postscript of SODA version]
[PDF of SICOMP version]
[PDF slides]
-
Pass Efficient Algorithms for Facility Location
Manuscript. Please note the number of the tech report.
[Postscript of Yale/DCS/TR-1337]
-
Multiple pass streaming algorithms for learning mixtures of
distributions in R^d
Theoretical Computer Science 410(19): 1765-1780 2009, special issue for ALT 2007. A preliminary version appeared in ALT 2007.
[Postscript of ALT version]
[PDF of TCS version]
[PDF slides]
-
Online and Offline Selling in Limit Order Markets
with Aaron Johson
in WINE 2008.
[Postcript of WINE version][postscript of full version]
-
An SDP Primal-Dual Algorithm for Approximating the Lovasz Theta Function
with Hubert Chan and Rajiv Raman
in ISIT 2009.
[PDF of ISIT version]
Check out my DBLP entry.
Miscellaneous
My Erdös number is 3.
Contact
E-mail: kchang at mpi-sb.mpg.de.
Telephone: +49 681 9325 113
Last modified October 12, 2007.