Sign Up to like & get
recommendations!
0
Published in 2017 at "IEEE Transactions on Information Theory"
DOI: 10.1109/tit.2017.2735799
Abstract: We study online learning under logarithmic loss with regular parametric models. In this setting, each strategy corresponds to a joint distribution on sequences. The minimax optimal strategy is the normalized maximum likelihood (NML) strategy. We…
read more here.
Keywords:
normalized maximum;
maximum likelihood;
characterizes optimality;
strategy ... See more keywords