A POMDP model which solves proactive learning -- a generalized realistic active learning scenario with multiple, imprecise, irresponsive, cost-varying oracles. This source code implements the original three algorithms (PAL #1-3) proposed by Donmez and Carbonell, as well as our POMDP proactive learner. It also contains code to run experiments on the UCI Adult and Spambase datasets, with five different types of oracle scenarios (Original #1-3 and Complex #1-2).
For more information, please see our AAAI 2016 paper:
Wray, Kyle H. and Zilberstein, Shlomo. "A POMDP Formulation of Proactive Learning." In Proceedings of the Thirtieth Conference on Artificial Intelligence (AAAI), Phoenix, AZ, USA, February 2016.