Computer Science – Learning
Scientific paper
2008-05-15
Computer Science
Learning
Scientific paper
We propose a framework for analyzing and comparing distributions, allowing us to design statistical tests to determine if two samples are drawn from different distributions. Our test statistic is the largest difference in expectations over functions in the unit ball of a reproducing kernel Hilbert space (RKHS). We present two tests based on large deviation bounds for the test statistic, while a third is based on the asymptotic distribution of this statistic. The test statistic can be computed in quadratic time, although efficient linear time approximations are available. Several classical metrics on distributions are recovered when the function space used to compute the difference in expectations is allowed to be more general (eg. a Banach space). We apply our two-sample tests to a variety of problems, including attribute matching for databases using the Hungarian marriage method, where they perform strongly. Excellent performance is also obtained when comparing distributions over graphs, for which these are the first such tests.
Borgwardt Karsten
Gretton Arthur
Rasch Malte J.
Schölkopf Bernhard
Smola Alexander J.
No associations
LandOfFree
A Kernel Method for the Two-Sample Problem does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with A Kernel Method for the Two-Sample Problem, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and A Kernel Method for the Two-Sample Problem will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-510669