A generic solution to Multi-Armed Bernoulli Bandit problems based on random sampling from sibling conjugate priors

ICAART 2010 - 2nd International Conference on Agents and Artificial Intelligence, Proceedings

Volumn 1, Issue , 2010, Pages 36-44

  Norheim, Thomas ^a   Brådland, Terje ^a   Granmo, Ole Christoffer ^a   Oommen, B John ^b  

^a UNIVERSITY OF AGDER   (Norway)

^b CARLETON UNIVERSITY   (Canada)

Author keywords

Bandit problems; Bayesian learning; Conjugate priors; Sampling

Indexed keywords

ABSTRACT FRAMEWORK; BANDIT PROBLEMS; BAYESIAN; BAYESIAN LEARNING; BAYESIAN METHODS; BERNOULLI; CLASSICAL OPTIMIZATION; CONJUGATE PRIOR; GENERIC SOLUTIONS; HYPER-PARAMETERS; MULTIPLE ARMS; NOVEL SOLUTIONS; OPTIMAL DECISION MAKING; RANDOM REWARD; RANDOM SAMPLING;

DECISION MAKING; MULTI AGENT SYSTEMS; OPTIMIZATION;

BAYESIAN NETWORKS;

EID: 77956297112     PISSN: None     EISSN: None     Source Type: Conference Proceeding    
DOI: None     Document Type: Conference Paper

References (0)