Balancing Exploitation and Exploration via Fully Probabilistic Design of Decision Policies

0495875 - ÚTIA 2019 CZ eng V - Výzkumná zpráva
Kárný, Miroslav - Hůla, František
Balancing Exploitation and Exploration via Fully Probabilistic Design of Decision Policies.
Praha: ÚTIA AV ČR, v.v.i, 2018. 13 s. Research Report, 2376.
Grant CEP: GA ČR GA16-09848S; GA ČR(CZ) GA18-15970S
Institucionální podpora: RVO:67985556
Klíčová slova: Exploitation * Exploration * Bayesian estimation * Adaptive systems * Fully probabilistic design * Kullback-Leibler divergence * Decision policy * Markov decision process
Obor OECD: Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
http://library.utia.cas.cz/separaty/2018/AS/karny-0495875.pdf

Adaptive decision making learns an environment model serving a design of a decision policy. The policy-generated actions influence both the acquired reward and the future knowledge. The optimal policy properly balances exploitation with exploration. The inherent dimensionality
curse of decision making under incomplete knowledge prevents the realisation of the optimal design.
Trvalý link: http://hdl.handle.net/11104/0288947