Optimal Reinforcement Learning with Black Holes
We introduce the Black Hole Reinforcement Learning problem, a previously unexplored variant of reinforcement learning in which we lose all turn information and all reward from trajectories that visit a particular subset of states. We assume awareness of the trajectory loss events, making this a cens...
Main Author: | Micali, Enrico |
---|---|
Other Authors: | Daskalakis, Constantinos |
Format: | Thesis |
Published: |
Massachusetts Institute of Technology
2022
|
Online Access: | https://hdl.handle.net/1721.1/143151 |
Similar Items
-
The black hole
by: Varuna Vyas
Published: (2021-01-01) -
Investigation of black hole complementarity in AdS2 black holes
by: Wontae Kim, et al.
Published: (2024-02-01) -
Causality constraints on black holes beyond GR
by: Francesco Serra, et al.
Published: (2022-08-01) -
White Hole-Black Hole Algorithm
by: Suad Khairi, Mohammed, et al.
Published: (2016) -
Black Holes and the loss landscape in machine learning
by: Pranav Kumar, et al.
Published: (2023-10-01)