Počet záznamů: 1
Approximate Dynamic Programming Based on High Dimensional Model Representation
- 1.
SYSNO ASEP 0399560 Druh ASEP J - Článek v odborném periodiku Zařazení RIV J - Článek v odborném periodiku Poddruh J Článek ve WOS Název Approximate Dynamic Programming Based on High Dimensional Model Representation Tvůrce(i) Pištěk, Miroslav (UTIA-B) RID, ORCID Zdroj.dok. Kybernetika. - : Ústav teorie informace a automatizace AV ČR, v. v. i. - ISSN 0023-5954
Roč. 49, č. 5 (2013), s. 720-737Poč.str. 18 s. Forma vydání Tištěná - P Jazyk dok. eng - angličtina Země vyd. CZ - Česká republika Klíč. slova approximate dynamic programming ; Bellman equation ; approximate HDMR minimization ; trust region problem Vědní obor RIV BC - Teorie a systémy řízení CEP GAP102/11/0437 GA ČR - Grantová agentura ČR Institucionální podpora UTIA-B - RVO:67985556 UT WOS 000328665200004 EID SCOPUS 84889006605 Anotace This article introduces an algorithm for implicit High Dimensional Model Representation (HDMR) of the Bellman equation. This approximation technique reduces memory demands of the algorithm considerably. Moreover, we show that HDMR enables fast approximate min- imization which is essential for evaluation of the Bellman function. In each time step, the problem of parametrized HDMR minimization is relaxed into trust region problems, all sharing the same matrix. Finding its eigenvalue decomposition, we effectively achieve estimates of all minima. Their full-domain representation is avoided by HDMR and then the same approach is used recursively in the next time step. An illustrative example of N-armed bandit problem is included. We assume that the newly established connection between approximate HDMR minimization and the trust region problem can be beneficial also to many other applications. Pracoviště Ústav teorie informace a automatizace Kontakt Markéta Votavová, votavova@utia.cas.cz, Tel.: 266 052 201. Rok sběru 2014
Počet záznamů: 1