Improved Path-length Regret Bounds for Bandits.

Sébastien Bubeck Yuanzhi Li Haipeng Luo Chen-Yu Wei

Published in: COLT (2019)

Keyphrases

path length
regret bounds
multi armed bandit
shortest path
online learning
lower bound
linear regression
upper bound
small world
learning algorithm
bayesian networks
social media
nearest neighbor
mutual information
em algorithm