Zobrazeno 1 - 6
of 6
pro vyhledávání: '"Norbert Kozlowski"'
Publikováno v:
IEEE Access, Vol 11, Pp 41190-41204 (2023)
Deep reinforcement learning with Experience Replay (ER), including Deep Q-Network (DQN), has been used to solve many multi-step learning problems. However, in practice, DQN algorithms need better explainability, which limits their applicability in ma
Externí odkaz:
https://doaj.org/article/dc2bd69884964b86b3b1256583fb2cf0
Autor:
Norbert Kozlowski, Olgierd Unold
Publikováno v:
IEEE Access, Vol 10, Pp 33816-33828 (2022)
Real-valued environments are challenging for learning systems because of a significant increase in the input space size of the problem. This work demonstrates that Anticipatory Learning Classifier Systems (ALCS) can successfully build sets of conditi
Externí odkaz:
https://doaj.org/article/ef44455a2f794c9eae42a13e0aa61bdf
Autor:
Norbert Kozlowski, Olgierd Unold
Publikováno v:
Applied Sciences, Vol 11, Iss 1098, p 1098 (2021)
Applied Sciences
Volume 11
Issue 3
Applied Sciences
Volume 11
Issue 3
Initially, Anticipatory Classifier Systems (ACS) were designed to address both single and multistep decision problems. In the latter case, the objective was to maximize the total discounted rewards, usually based on Q-learning algorithms. Studies on
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::278c3f370339b951685d3402fee17244
https://www.mdpi.com/2076-3417/11/3/1098
https://www.mdpi.com/2076-3417/11/3/1098
Autor:
Norbert Kozlowski, Olgierd Unold
Publikováno v:
GECCO Companion
One way of dealing with the real-valued input signal is to discretize it. This might influence the process of learning the environmental model by the ACS2 agent. A more sophisticated method of selecting action can be applied to increase the speed of
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::47a413e0d7fb3ef4b5dcca6236970d9f
https://doi.org/10.1145/3377929.3398079
https://doi.org/10.1145/3377929.3398079
Publikováno v:
Advances in Intelligent Systems and Computing ISBN: 9783030197377
CORES
CORES
This paper introduces and tests Action Planning mechanism in the Anticipatory Classifier System ACS2. Action Planning implies goal-directed learning and bidirectional search to strengthen reliable classifiers. It is shown that it can speed up the pro
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::f45800d015e6c5981981af1386fb4535
https://doi.org/10.1007/978-3-030-19738-4_27
https://doi.org/10.1007/978-3-030-19738-4_27
Autor:
Olgierd Unold, Norbert Kozlowski
Publikováno v:
GECCO (Companion)
This paper explains the process of integrating ACS2 algorithm with the standardised framework for comparing reinforcement learning tasks - OpenAI Gym. The new Python library is introduced alongside with standard environments derived from LCS literatu
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::42adf2a8c01bb6943c1267add25ebb5f
https://doi.org/10.1145/3205651.3208241
https://doi.org/10.1145/3205651.3208241