BrainTube » mean http://braintube.tv Braincandy for your needs... Tue, 22 Jun 2010 23:58:10 +0000 http://wordpress.org/?v=2.8.4 en hourly 1 Countdown to MIX10: Drinking Las Vegas http://channel9.msdn.com/posts/LarryLarsen/Countdown-to-MIX10-Drinking-Las-Vegas/ http://channel9.msdn.com/posts/LarryLarsen/Countdown-to-MIX10-Drinking-Las-Vegas/#comments Thu, 04 Mar 2010 06:41:00 +0000 channel9 http://ecn.channel9.msdn.com/o9/ch9/1/1/0/4/3/5/FinalMix10Show_ch9.wmv

Sure, sure. MIX is mostly all about the great keynotes, sessions, networking, and all the awesome content that is going to make your brain hurt in a good way, but that doesn’t mean you can’t go out for a good time while you’re in Sin City! On this, our last Countdown to MIX10 show, @anyware and @ritzy get advice from THE cocktail evangelist, Robert Hess, who tells us where to find a mean cocktail in Vegas. If we’re not at the conference, you might find us bellied up to a bar somewhere, and if you do, maybe your next round will be on us.  See ya in Vegas!

This is our last show for MIX! Thanks all for your tremendous help!!!

]]>

Sure, sure. MIX is mostly all about the great keynotes, sessions, networking, and all the awesome content that is going to make your brain hurt in a good way, but that doesn’t mean you can’t go out for a good time while you’re in Sin City! On this, our last Countdown to MIX10 show, @anyware and @ritzy get advice from THE cocktail evangelist, Robert Hess, who tells us where to find a mean cocktail in Vegas. If we’re not at the conference, you might find us bellied up to a bar somewhere, and if you do, maybe your next round will be on us.  See ya in Vegas!

This is our last show for MIX! Thanks all for your tremendous help!!!

]]> http://channel9.msdn.com/posts/LarryLarsen/Countdown-to-MIX10-Drinking-Las-Vegas/feed/ 0 The Query Complexity of Estimating Weighted Averages http://www.youtube.com/watch?v=qyxIUPTeNfI&feature=youtube_gdata http://www.youtube.com/watch?v=qyxIUPTeNfI&feature=youtube_gdata#comments Tue, 16 Feb 2010 23:24:46 +0000 Uploads by googletechtalks http://gdata.youtube.com/feeds/base/videos/qyxIUPTeNfI
Google Tech Talk February 4, 2010 ABSTRACT Presented by Tony Wirth. The query complexity of estimating the mean of some [0, 1] variables is well known to the theory community. Inspired by some work by Carterette et al. [SIGIR 2006, pp 26875] on evaluating retrieval systems, and by Moffat and Zobel's new proposal for such evaluation [under review], we decided to examine the query complexity of weighted average calculation. In general, the problem requires the same number of queries as estimating the mean, as the latter is a special case. In fact, there is a matching upper bound for the weighted mean. This result remains true for any set of weights that is the normalized prefix of a divergent series. However, if the weights follow a geometric sequence, a much smaller sample is sufficient. Finally, we investigate power-law sequences of weights and show matching lower and upper bounds. This is joint work with Amit Chakrabarti and Venkatesan Guruswami and Andrew Wirth. Tony Wirth joined the faculty of the University of Melbourne's Computer Science department in 2005. Prior to that, he completed a PhD, as a Gordon Wu Fellow, at Princeton University in 2004 on approximation algorithms for clustering problems. Tony completed his undergraduate degree at the University of Melbourne, majoring in statistics. His research interests also include sequence problems in bioinformatics and adaptive sampling.
Views: 1579
5 ratings
Time: 41:30 More in Science & Technology]]>
Google Tech Talk February 4, 2010 ABSTRACT Presented by Tony Wirth. The query complexity of estimating the mean of some [0, 1] variables is well known to the theory community. Inspired by some work by Carterette et al. [SIGIR 2006, pp 26875] on evaluating retrieval systems, and by Moffat and Zobel's new proposal for such evaluation [under review], we decided to examine the query complexity of weighted average calculation. In general, the problem requires the same number of queries as estimating the mean, as the latter is a special case. In fact, there is a matching upper bound for the weighted mean. This result remains true for any set of weights that is the normalized prefix of a divergent series. However, if the weights follow a geometric sequence, a much smaller sample is sufficient. Finally, we investigate power-law sequences of weights and show matching lower and upper bounds. This is joint work with Amit Chakrabarti and Venkatesan Guruswami and Andrew Wirth. Tony Wirth joined the faculty of the University of Melbourne's Computer Science department in 2005. Prior to that, he completed a PhD, as a Gordon Wu Fellow, at Princeton University in 2004 on approximation algorithms for clustering problems. Tony completed his undergraduate degree at the University of Melbourne, majoring in statistics. His research interests also include sequence problems in bioinformatics and adaptive sampling.
Views:
1579
5
ratings

Time:
41:30
More in
Science & Technology ]]> http://www.youtube.com/watch?v=qyxIUPTeNfI&feature=youtube_gdata/feed/ 0