prikaz prve stranice dokumenta Implementacija i validacija algoritma za poticano strojno učenje
No public access
undergraduate thesis
Implementacija i validacija algoritma za poticano strojno učenje
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2024. urn:nbn:hr:168:474217
Mikan, Bruno
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Mikan, B. (2024). Implementacija i validacija algoritma za poticano strojno učenje (Undergraduate thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:474217

Mikan, Bruno. "Implementacija i validacija algoritma za poticano strojno učenje." Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2024. https://urn.nsk.hr/urn:nbn:hr:168:474217

Mikan, Bruno. "Implementacija i validacija algoritma za poticano strojno učenje." Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2024. https://urn.nsk.hr/urn:nbn:hr:168:474217

Mikan, B. (2024). 'Implementacija i validacija algoritma za poticano strojno učenje', Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 27 November 2024, https://urn.nsk.hr/urn:nbn:hr:168:474217

Mikan B. Implementacija i validacija algoritma za poticano strojno učenje [Undergraduate thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2024 [cited 2024 November 27] Available at: https://urn.nsk.hr/urn:nbn:hr:168:474217

B. Mikan, "Implementacija i validacija algoritma za poticano strojno učenje", Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2024. Available at: https://urn.nsk.hr/urn:nbn:hr:168:474217

Metadata
TitleImplementacija i validacija algoritma za poticano strojno učenje
Title (english)Implementation and Validation of Reinforcement Learning Algorithms
AuthorBruno Mikan
MentorDamir Pintar (mentor)
Committee memberDamir Pintar (predsjednik povjerenstva)
Committee memberLuka Humski (član povjerenstva)
Committee memberAna Šarčević (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2024-07-01, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Poticano učenje postaje sve važnija grana umjetne inteligencije te se koristi za rješavanje sve složenijih problema. Matematička osnova poticanog učenja formalno je opisana konačnim Markovljevim procesima odlučivanja te smjerova za daljnji razvoj i istraživanja ima puno. U ovom radu dan je pregled osnove teorijske podloge poticanog učenja, implementirana je vlastita okolina s konačnim i diskretnim skupom stanja i akcija na kojoj se može trenirati agent poticanog učenja. Također, opisana je vlastita implementacija algoritma Q-učenja te je uspješno provedena usporedba s kontrolnim agentom i već gotovim agentom. Usporedbom je dokazana uspješnost implementacije algoritma Q-učenja na problemu igre platforme i loptice. Konačno, dan je pregled mogućih smjerova daljnjeg razvoja okoline.
Abstract (english)
Reinforcement learning is becoming an increasingly important branch of artificial intelligence and is used to solve increasingly complex problems. The mathematical basis of reinforcement learning is formally described by the finite Markov decision process, and there are many directions for further development and research. In this work, an overview of the theoretical basis of supported learning is given. Also, an own environment with a finite and discrete state set and action set is implemented on which a reinforcement learning agent can be trained. Also, the implementation of the Q-learning algorithm was described, and a comparison was successfully carried out with a control agent and an already finished agent. The comparison proved the success of the implementation of the Q-learning algorithm on the platform and ball game problem. Finally, an overview of some possible directions for further development of the environment is given.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:474217
Study programmeTitle: Computing
Study programme type: university
Study level: undergraduate
Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Public note
Created on2024-10-12 21:35:01