PUBLICATIONS (Reverse order)

Presentations:-

4	Avrim Blum, T.-H. Hubert Chan, Mugizi Robert Rwebangira, “A Random-Surfer Web-Graph Model”, ANALCO '06. (ppt) I gave this presentation on March 27, 2006 at the Theory lunch at CMU. It is essentially an expanded version of the ANALCO ‘06 talk.
3	Avrim Blum, T.-H. Hubert Chan, Mugizi Robert Rwebangira, “A Random-Surfer Web-Graph Model”, ANALCO '06. (ppt) This was the presentation I gave at the ANALCO ’06 workshop on January 21, 2006 at the Radisson Hotel in Miami.
2	A. Blum, M.R. Rwebangira, R. Reddy, J. Lafferty “Improving the Graph Mincut Approach to Learning from Labeled and Unlabeled Examples” 2003 (ppt) I presented my ICML poster (with a few changes) to the Machine Learning lunch at CMU in September 2003.
1	A. Blum, M.R. Rwebangira, R. Reddy, J. Lafferty “Improving the Graph Mincut Approach to Learning from Labeled and Unlabeled Examples”, International Conference on Machine Learning 2003 (Non refereed Poster) (ppt) This was a poster I presented about work in progress at ICML in August 2003.

Publications:-

5	Avrim Blum, T.-H. Hubert Chan, Mugizi Robert Rwebangira, “A Random-Surfer Web-Graph Model”, ANALCO '06. (pdf) (ppt1) (ppt2) This work was done in Fall 2005. We studied a certain natural model for producing a random graph and did some experimental and theoretical analysis. I later presented this at ANALCO in Miami and at the Theory lunch at CMU.
4	Maria-Florina Balcan, Avrim Blum, Pakyan Choi, John Lafferty, Brian Pantano, Mugizi Robert Rwebangira, Xiaojin Zhu “Person identification in webcam images: An application of semi-supervised learning”, ICML 2005 Workshop on Learning with Partially Classified Training Data (pdf) This work was done during 2004 and 2005 by a large group of collaborators. It was essentially a large scale application of semi-supervised learning to a “realistic” task. Jerry Zhu was the lead and presented it at a workshop in ICML 2005.
3	A. Blum, J. Lafferty, M.R. Rwebangira, R. Reddy “Semi-supervised Learning Using Randomized Mincuts”, International Conference on Machine Learning 2004 (ps) (pdf) This is an extension of the graph mincut approach to learning with labeled and unlabeled data originally proposed by Blum and Chawla. We added randomness to the graph structure and took the average of several mincuts. We also proposed a general method for constructing the graph that seems to have robust performance.
2	D. Sow, G. Banavar, J.S. Davis II, J Sussman, M.R. Rwebangira “Preparing the Edge of the Network for Pervasive Content Delivery”, Advanced Topic Workshop on Middleware for Mobile Computing with IFIP/ACM middleware conference 2001 (pdf) This work was done in the summer of 2001 at IBM research and published in fall of the same year. The title pretty much says it all; it’s all about pervasive computing.
1	L. Burge, M.R. Rwebangira. Constructing Reliable Software across the ORB. Symposium on Computing at Minority Institutions. ADMI 2000. (html)