4 ms·
Fast Ranking Algorithm: Astonishing Paper by Raykar, Duraiswami, and Krishnapuram
- aswanson 18y agoI suppose you have an IEEE $ub$cription or member$hip?
- startingup 18y agoPDF here: http://www.umiacs.umd.edu/~vikas/publications/raykar_AISTATS_2007.pdf http://www.umiacs.umd.edu/~vikas/publications/raykar_AISTATS...
- greendestiny 18y agoI got this link, its probably the same: http://www.umiacs.umd.edu/~vikas/publications/raykar_PAMI_2007.pdf http://www.umiacs.umd.edu/~vikas/publications/raykar_PAMI_20...
- startingup 18y agoPaper from Google Research in a related area: http://www.cs.nyu.edu/web/Research/TechReports/TR2007-903/TR2007-903.pdf http://www.cs.nyu.edu/web/Research/TechReports/TR2007-903/TR...
- natch 18y agoIf an order of magnitude speedup is "astonishing" then the field must have become overconfident and complacent. Great to see someone rocking the boat. But does anyone care to explain any more (in English, for us plebes) more about how this is astonishing? I'm not challenging the assessment; I'm just curious about what it means. And some interesting words from one of the other papers: what are "classifier regret" and "ranking regret"? Is that a metric of how often you go down the wrong path or something?
- melvinram 18y agoDoesn't directly answer your question, but read the comment by thedelimiter on the page below and it may provide some insight :D http://www.reddit.com/info/6nh28/comments/c04d5l0 http://www.reddit.com/info/6nh28/comments/c04d5l0
- natch 18y agohehe, yes, that was well done.
- brent 18y agoAre you saying an improvement from O(n^2) to O(n) is not impressive? BTW, if this is the improvement you are referring to, its not a "order of magnitude".