Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/150645
Title: Aplicació d’algoritmes d’intel·ligència artificial per al desenvolupament d’agents per als jocs de cartes “Brisca” i “Tute”
Author: Arjonilla Zamora, Rubén
Tutor: Sanchez, Friman  
Nunez do Rio, Joan M  
Others: Acedo Nadal, Susana
Abstract: Brisca and Tute are two strategic card games with imperfect information that have not been studied in the computational learning field, but they are quite interesting and have special interest in machine learning field for their strategic base. This research focuses on the development of agents capable of learning to play both games using supervised learning techniques (neural networks), genetic algorithms and reinforcement learning, and it covers from the environment and graphical user interface development to the training of models using different approaches for each technique. Models and results are detailed along a qualitatvie analysis of the learning capability of each technique and the models' performance playing the games. The strenghts and weaknesses of each technique and approachment are discussed, the limitations found during the training are detailed, and possible areas of improvement and future research directions are identified to continue advancing in computational learning applied to strategic card games with imperfect information. At the end of this research, experienced players have faced the different obtained models and won almost all matches against reinforcement and genetic trained agents, but it was more balanced when they faced supervised models. The obtained models still have a lot of room to improve and there is much more to investigate in this field.
Keywords: aprenentatge computacional
agent IA
cartes
brisca
tute
Document type: info:eu-repo/semantics/bachelorThesis
Issue Date: 16-Jun-2024
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
rarjonillaTFG0624memoria.pdfMemòria del TFG2,54 MBAdobe PDFThumbnail
View/Open
Share:
Export:
View statistics

This item is licensed under aCreative Commons License Creative Commons