Worst case prediction over sequences under log loss

MaRDI QIDQ4247216zbMATH OpenFDO

Authors David Haussler, Manfred Opper

Publication date 6 April 2000

zbMATH Keywords

logarithmic lossmetric entropiesminimax regretpredictionstatistical estimation

Mathematics Subject Classification ID

Statistical aspects of information-theoretic topics (62B10) Inference from stochastic processes and prediction (62M20) Applications of game theory (91A80) Measures of information, entropy (94A17)

Recommendations

• Worst-case bounds for the logarithmic loss of predictors
• On prediction of individual sequences
• Asymptotic minimax regret for data compression, gambling, and prediction

Cited in

(11)

• Predicting a binary sequence almost as well as the optimal biased coin
• On prediction of individual sequences
• Prediction in the worst case
• The fundamental nature of the log loss function
• Asymptotic minimax regret for data compression, gambling, and prediction
• scientific article; zbMATH DE number 7014218 (Why is no real title available?)
• Extended stochastic complexity and minimax relative loss analysis
• Worst-case bounds for the logarithmic loss of predictors
• Minimax rates for conditional density estimation via empirical entropy
• A tight excess risk bound via a unified PAC-Bayesian-Rademacher-Shtarkov-MDL complexity
• Sensor networks: from dependence analysis via matroid bases to online synthesis

This page was built for publication: Worst case prediction over sequences under log loss

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4247216)