CMU-ML-10-105
Machine Learning Department
School of Computer Science, Carnegie Mellon University


CMU-ML-10-105

The Sample Complexity of Self-Verifying
Bayesian Active Learning

Liu Yang, Steve Hanneke, Jaime Carbonell

June 2010

CMU-ML-10-105.pdf


Keywords: Bayesian Active Learning, Learning Theory, Sample Complexity


We prove that access to the prior distribution over target functions can improve the sample complexity of self-terminating active learning algorithms, so that it is always better than the known results for prior-dependent passive learning. In particular, this is in stark contrast to the analysis of prior-independent algorithms, where there are known simple learning problems for which no self-terminating algorithm can provide this guarantee for all priors.

14 pages

SCS Technical Report Collection
School of Computer Science homepage

This page maintained by [email protected]