Online calibrated forecasts: memory efficiency versus universality for learning in games

DOI10.1007/S10994-006-0219-YzbMATH Open1471.91051OpenAlexW2037905382MaRDI QIDQ2384142FDOQ2384142

Authors: Shie Mannor, Jeff S. Shamma, Gürdal Arslan 

Publication date: 20 September 2007

Published in: Machine Learning (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1007/s10994-006-0219-y

Recommendations

• Learning in games with bounded memory
• Learning with bounded memory in games
• Prediction, Optimization, and Learning in Repeated Games
• Learning and equilibrium as useful approximations: accuracy of prediction on randomly selected constant sum games
• scientific article; zbMATH DE number 1804097
• Asymptotically optimal strategies for online prediction with history-dependent experts
• Smooth calibration, leaky forecasts, finite recall, and Nash dynamics

zbMATH Keywords

forecastingstochastic approximationcalibrationfictitious playlearning in gamesODE methodprediction of universal sequences

Mathematics Subject Classification ID

Rationality and learning in game theory (91A26)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• COMPLEXITY AND REAL COMPUTATION: A MANIFESTO
• Title not available (Why is that?)
• Evolutionary Games and Population Dynamics
• The Nonstochastic Multiarmed Bandit Problem
• Title not available (Why is that?)
• Calibrated learning and correlated equilibrium
• Title not available (Why is that?)
• Title not available (Why is that?)
• Adaptive Heuristics
• A general class of adaptive strategies
• Title not available (Why is that?)
• A separation principle for the control of a class of nonlinear systems
• Asymptotic calibration
• Title not available (Why is that?)
• Regret in the on-line decision problem
• Asynchronous stochastic approximation and Q-learning
• Stochastic approximation with two time scales
• The O.D.E. Method for Convergence of Stochastic Approximation and Reinforcement Learning
• Mixed equilibria and dynamical systems arising from fictitious play in perturbed games
• Dynamic fictitious play, dynamic gradient play, and distributed convergence to Nash equilibria
• Stochastic uncoupled dynamics and Nash equilibrium
• A game of prediction with expert advice
• Learning Theory
• Convergence rate of linear two-time-scale stochastic approximation.
• Calibration with Many Checking Rules
• An easier way to calibrate.
• Self-Calibrating Priors Do Not Exist
• Convergent multiple-timescales reinforcement learning algorithms in normal form games
• Equivalent necessary and sufficient conditions on noise sequences for stochastic approximation algorithms

Cited In (2)

• Deterministic calibration and Nash equilibrium
• Regret minimization in repeated matrix games with variable stage duration

This page was built for publication: Online calibrated forecasts: memory efficiency versus universality for learning in games

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2384142)