Finite-time Analysis of the Multiarmed Bandit Problem
Machine Learning
Stability of transductive regression algorithms
Proceedings of the 25th international conference on Machine learning
| Hi-index | 0.00 |
Active search is an increasingly important learning problem in which we use a limited budget of label queries to discover as many members of a certain class as possible. Numerous real-world applications may be approached in this manner, including fraud detection, product recommendation, and drug discovery. Active search has model learning and exploration/exploitation features similar to those encountered in active learning and bandit problems, but algorithms for those problems do not fit active search. Previous work on the active search problem [5] showed that the optimal algorithm requires a lookahead evaluation of expected utility that is exponential in the number of selections to be made and proposed a truncated lookahead heuristic. Inspired by the success of myopic methods for active learning and bandit problems, we propose a myopic method for active search on graphs. We suggest selecting points by maximizing a score considering the potential impact of selecting a node, meant to emulate lookahead while avoiding exponential search. We test the proposed algorithm empirically on real-world graphs and show that it outperforms popular approaches for active learning and bandit problems as well as truncated lookahead of a few steps.