New answers tagged multi-armed-bandits
1
vote
Accepted
UCB, Thompson sampling etc seems myopic/greedy for bandits?
The (binary) multi-armed bandit actually is a MDP with one state and $K$ actions.
So your suggestion boils down to meta-learning: Find the parameters of one MDP that can solve another. Let's go with ...
Top 50 recent answers are included
Related Tags
multi-armed-bandits × 60reinforcement-learning × 52
contextual-bandits × 16
upper-confidence-bound × 10
markov-decision-process × 8
thompson-sampling × 6
comparison × 4
policy-gradients × 4
sutton-barto × 4
epsilon-greedy-policy × 4
machine-learning × 3
reference-request × 3
terminology × 3
q-learning × 3
rewards × 3
deep-rl × 2
papers × 2
math × 2
definitions × 2
optimization × 2
time-series × 2
policies × 2
notation × 2
continuous-action-spaces × 2
discount-factor × 2