Uporaba drevesnega preiskovanja Monte Carlo za inteligentno vodenje igralnih agentov v Rogue-podobnih igrah : diplomsko delo

Jeran, Benjamin

| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

First page > Show document

Show document

Title:	Uporaba drevesnega preiskovanja Monte Carlo za inteligentno vodenje igralnih agentov v Rogue-podobnih igrah : diplomsko delo
Authors:	ID Jeran, Benjamin (Author) ID Fister, Iztok (Mentor) More about this mentor... ID Novak, Damijan (Comentor)
Files:	VS_Jeran_Benjamin_2024.pdf (2,00 MB) MD5: 5961F60336D8C67A0CAA9AB01DDA88D8
Language:	Slovenian
Work type:	Bachelor thesis/paper
Typology:	2.11 - Undergraduate Thesis
Organization:	FERI - Faculty of Electrical Engineering and Computer Science
Abstract:	V diplomski nalogi smo se osredotočili na implementacijo igralnega agenta za Rogue-podobne igre. V teoretičnem delu smo najprej opisali, kaj so Rogue-podobne igre ter izpostavili njihove glavne značilnosti. Naredili smo pregled igralnih agentov in njihovih pristopov ter se na koncu dotaknili osnovne implementacije algoritma drevesnega preiskovanja Monte Carlo, pri katerem smo uporabili formulo UCT za izbiranje vozlišč. V praktičnem delu smo s pomočjo algoritma zgradili igralnega agenta za Rogue-podobno igro ter naredili raziskavo, kako dobro se je ta obnesel v igranju. Raziskovali smo, kako dobro se agent na podlagi algoritma MCTS izkaže proti igranju nasprotnika, ki izbira naključne poteze. Algoritem MCTS se je ob koncu praktičnega dela dobro izkazal za uporabo v Rogue-podobnih igrah.
Keywords:	Rogue-podobne igre, MCTS, inteligentno vodenje agentov, okrepitveno učenje
Place of publishing:	Maribor
Place of performance:	Maribor
Publisher:	[B. Jeran]
Year of publishing:	2024
Number of pages:	1 spletni vir (1 datoteka PDF (X, 52 f.))
PID:	20.500.12556/DKUM-90098
UDC:	004.832.24(043.2)
COBISS.SI-ID:	218920451
Publication date in DKUM:	08.10.2024
Views:	141
Downloads:	111
Metadata:
Categories:	KTFMB - FERI
:	Copy citation

Average score:	(0 votes)
Your score:	Voting is allowed only for logged in users.
Share:

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:	CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Link:	http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:	The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:	22.08.2024

Secondary language

Language:	English
Title:	Using the Monte Carlo tree search algorithm for intelligent guiding of game agents in rogue-like games
Abstract:	In the thesis, the focus was placed on the implementation of a game-playing agent for Rogue-like games. In the theoretical part, Rogue-like games were first described, and their main characteristics were highlighted. A review of game-playing agents and their approaches was conducted, and finally, the basic implementation of the Monte Carlo tree search algorithm was discussed, where the UCT formula was used for node selection. In the practical section, the algorithm was used to create a game-playing agent for a Rogue-like game, and its gameplay performance was evaluated through analysis. Research was conducted on the performance of an agent based on the MCTS algorithm against an opponent that makes random moves. By the end of the practical part, the MCTS algorithm proved to be effective for use in Rogue-like games.
Keywords:	Roguelike, MCTS, intelligent guiding of game agents, reinforcement learning

Comments

Leave comment

You must log in to leave a comment.

Comments (0)

0 - 0 / 0

There are no comments!

Back