Learning from Multiple Sources
Koby Crammer, Michael J. Kearns, Jennifer Russo Wortman · The MIT Press eBooks · 2007
Abstract We consider the problem of learning accurate models from multiple sources of"nearby " data. Given distinct samples from multiple data sources and estimates of the dissimilarities between these sources, we provide a general theory of whichsamples should be used to learn models for each source. This theory is applicable in a broad decision-theoretic learning framework, and yields results for classifi-cation and regression generally, and for density estimation within the exponential family. A key component of our approach is the development of approximatetriangle inequalities for expected loss, which may be of independent interest.