Temporal Difference Uncertainties as a Signal for Exploration

Sebastian Flennerhag
Sebastian Flennerhag
Francesco Visin
Francesco Visin
Alexandre Galashov
Alexandre Galashov
Diana L. Borsa
Diana L. Borsa
Andre Barreto
Andre Barreto
Cited by: 0|Bibtex|Views24|Links

Abstract:

An effective approach to exploration in reinforcement learning is to rely on an agent's uncertainty over the optimal policy, which can yield near-optimal exploration strategies in tabular settings. However, in non-tabular settings that involve function approximators, obtaining accurate uncertainty estimates is almost as challenging a pr...More

Code:

Data:

Your rating :
0

 

Tags
Comments