Back to search

Bc. Richard Schwarz

Bachelor's thesis

Monte Carlo Tree Search in Deep Reinforcement Learning Algorithms

Monte Carlo Tree Search in Deep Reinforcement Learning Algorithms
Abstract:
Táto práca skúma integráciu Monte Carlo stromového vyhľadávania (MCTS) do algoritmov hlbokého posilňovacieho učenia. Po prvé, predstavujeme MCTS ako samostatnú politiku pre Markovské rozhodovacie procesy (MDP). Po druhé, kombinujeme ho s prístupmi založenými na modelovom posilňovacom učení, pričom využívame MCTS ako plánovací nástroj. Začíname s AlphaZero, ktorý operuje pod silnými predpokladmi o znalostiach …more
Abstract:
This thesis explores the integration of Monte Carlo tree search (MCTS) into deep reinforcement learning algorithms. Firstly, we introduce MCTS as a standalone policy for Markov decision processes (MDP). Secondly, we combine it with model-based reinforcement learning approaches by utilizing MCTS as a planning tool. We start with AlphaZero, which operates under strong assumptions about the knowledge …more
 

Keywords

Monte Carlo tree search deep reinforcement learning model-based reinforcement learning UCB AlphaZero MuZero Stochastic MuZero neural network
 
Language used: English
Date on which the thesis was submitted / produced: 23. 5. 2024
Identifier: https://is.muni.cz/th/k37az/

Thesis defence

  • Date of defence: 24. 6. 2024
  • Supervisor: doc. RNDr. Petr Novotný, Ph.D.
  • Reader: Mgr. Martin Kurečka

Citation record

Cite this text

Full text of thesis

Contents of on-line thesis archive
Published in Theses:
  • světu
Other ways of accessing the text
Institution archiving the thesis and making it accessible: Masarykova univerzita, Fakulta informatiky
Složka Reference to the local database directory of the institution