A Competitive Algorithm for Agnostic Active Learning
Abstract
For some hypothesis classes and input distributions, active agnostic learning needs exponentially fewer samples than passive learning; for other classes and distributions, it offers little to no improvement. The most popular algorithms for agnostic active learning express their performance in terms of a parameter called the disagreement coefficient, but it is known that these algorithms are inefficient on some inputs. We take a different approach to agnostic active learning, getting an algorithm that is competitive with the optimal algorithm for any binary hypothesis class and distribution over . In particular, if any algorithm can use queries to get error, then our algorithm uses queries to get error. Our algorithm lies in the vein of the splitting-based approach of Dasgupta [2004], which gets a similar result for the realizable () setting. We also show that it is NP-hard to do better than our algorithm's overhead in general.
Cite
@article{arxiv.2310.18786,
title = {A Competitive Algorithm for Agnostic Active Learning},
author = {Eric Price and Yihan Zhou},
journal= {arXiv preprint arXiv:2310.18786},
year = {2024}
}