Sequential prediction of individual sequences under general loss functions

DOI10.1109/18.705569zbMath1026.68579MaRDI QIDQ4701166

David Haussler, Manfred K. Warmuth, Jyrki Kivinen

Publication date: 21 November 1999

Published in: IEEE Transactions on Information Theory (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1109/18.705569

zbMATH Keywords

data compression; binary sequences; forecasting; regret; on-line learning

Mathematics Subject Classification ID

68T05: Learning and adaptive systems in artificial intelligence

94A05: Communication theory

Related Items

Unnamed Item, Competitive On-line Statistics, Regression Estimation from an Individual Stable Sequence, Probability theory for the Brier game, Optimal learning with Bernstein Online Aggregation, Sparse regression learning by aggregation and Langevin Monte-Carlo, Mirror averaging with sparsity priors, Generalized mirror averaging and \(D\)-convex aggregation, Robust forecast combinations, Sensor networks: from dependence analysis via matroid bases to online synthesis, Learning by mirror averaging, The weak aggregating algorithm and weak mixability, Supermartingales in prediction with expert advice, Efficient learning with virtual threshold gates, Loss functions, complexities, and the Legendre transformation., Universal prediction of random binary sequences in a noisy environment, Does snooping help?, Randomized prediction of individual sequences, Fast learning rates in statistical inference through aggregation, How many strings are easy to predict?, The Fundamental Nature of the Log Loss Function, Prediction with Expert Evaluators’ Advice