Online Non-Additive Path Learning under Full and Partial Information
algorithmic learning theory, 2019.
We consider the online path learning problem in a graph with non-additive gains/losses. Various settings of full information, semi-bandit, and full bandit are explored. We give an efficient implementation of EXP3 algorithm for the full bandit setting with any (non-additive) gain. Then, focusing on the large family of non-additive count-ba...More
PPT (Upload PPT)