Technical Program

Paper Detail

Paper: PS-2B.53
Session: Poster Session 2B
Location: H Fl├Ąche 1.OG
Session Time: Sunday, September 15, 17:15 - 20:15
Presentation Time:Sunday, September 15, 17:15 - 20:15
Presentation: Poster
Publication: 2019 Conference on Cognitive Computational Neuroscience, 13-16 September 2019, Berlin, Germany
Paper Title: A way around the exploration-exploitation dilemma
Manuscript:  Click here to view manuscript
License: Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
DOI: https://doi.org/10.32470/CCN.2019.1365-0
Authors: Erik Peterson, Timothy Verstynen, Carnegie Mellon, United States
Abstract: The exploration-exploitation dilemma is considered a fundamental but intractable problem in the learning and decision sciences. At its crux is the search to maximize reward. Here we challenge this view and show a way around the dilemma by defining separate mathematical objectives for exploration and exploitation. To make the objective for exploration independent of reward, we derive a set of general axioms for information value. Using these axioms we develop a greedy algorithm which provably and optimally maximizes both information and reward.