Bandit Regret Scaling with the Effective Loss Range.

Nicolò Cesa-Bianchi Ohad Shamir

Published in: ALT (2018)

Keyphrases

regret bounds
wide range
data sets
data mining
real time
neural network
website
case study
high quality
training set
dynamic programming
online learning
markov chain
bandit problems
upper confidence bound
multi armed bandit problems