Tuesday, November 25, 2008

Conference slides..

Just got back from conference, and here are the slides from given talk. All in all - not bad. Now - off to the further research on dimension reduction using query-induced subgraphs. Got really interesting question at the conference - what is the minimal number of nodes for which the similarity of \beta (power-law exponent) for entire graph and of induced subgraph holds? This would be a "phase transition" figure bellow for which \beta should explode, and all nice properties (existence of 'giant' component, walk convergence, etc) - should disappear. Anyhow ... we'll see :) . Meanwhile - you can checkout the paper - here . Stay tuned. ;)


Post a Comment

<< Home