On asymptotically optimal methods of prediction and adaptive coding for Markov sources
Journal of Complexity
How to Achieve Minimax Expected Kullback-Leibler Distance from an Unknown Finite Distribution
ALT '02 Proceedings of the 13th International Conference on Algorithmic Learning Theory
Bernstein polynomials and learning theory
Journal of Approximation Theory
k-norm misclassification rate estimation for decision trees
ASC '07 Proceedings of The Eleventh IASTED International Conference on Artificial Intelligence and Soft Computing
PAC-Bayesian Analysis of Co-clustering and Beyond
The Journal of Machine Learning Research
AusDM '12 Proceedings of the Tenth Australasian Data Mining Conference - Volume 134
Hi-index | 754.84 |
This correspondence shows that the known “add-half” rule is not asymptotically optimal for predicting the (n+1)st symbol after a sequence of n symbols, whereas the “add-β0” rule, β0=0.50922···is