Institutional Repository of Vilnius University
Q-mokymo algoritmų taikymas uždaviniams su diskrečiomis būsenų ir veiksmų aibėmis /
Abstract
dc:descriptionThe paper analyzes one of the main reinforcement learning methods - Q-learning, which is actively used to solve problem when output is unknown. Additionally, a dynamic Q-table algorithm is proposed to solve the main Q-learning drawback - its difficult practical applicability in environments with continues state spaces. Aim of this work is to implement a dynamic Q-table method, compare it to Q-learning and deep Q-learning using Python programming language with Open AI Gym Frozen Lake and CartPole environments.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2021
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Novikevičius, Edvardas,
- Contributors dc:contributor
-
- Kybartas, Rimantas
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD107116907&prefLang=en_US