Soft-Bayes: prod for mixtures of experts with log-loss

MaRDI QIDQ4645656zbMATH OpenFDO

Authors Laurent Orseau, Tor Lattimore, Shane Legg

Publication date 10 January 2019

Full work available at URL https://arxiv.org/abs/1901.02230, http://proceedings.mlr.press/v76/orseau17a.html

zbMATH Keywords

prediction with expert adviceonline convex optimizationlogarithmic loss

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Convex programming (90C25) Online algorithms; streaming algorithms (68W27)

Recommendations

• How to better use expert advice
• Algorithmic Learning Theory
• 10.1162/153244303321897654
• How to use expert advice
• Learning Theory

Cited in

(4)

• Algorithmic Learning Theory
• Mixability is Bayes risk curvature relative to log loss
• Market selection and learning under model misspecification
• A general equilibrium model of investor sentiment

This page was built for publication: Soft-Bayes: prod for mixtures of experts with log-loss

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4645656)