Neural networks applied to a tower defense video game
Metadatos
Mostrar el registro completo del ítemcomunitat-uji-handle:10234/158176
comunitat-uji-handle2:10234/71324
comunitat-uji-handle3:10234/169451
comunitat-uji-handle4:
TFG-TFMMetadatos
Título
Neural networks applied to a tower defense video gameAutoría
Tutor/Supervisor; Universidad.Departamento
Montoliu Colás, Raul; Universitat Jaume I. Departament d'Enginyeria i Ciència dels ComputadorsFecha de publicación
2018-06Editor
Universitat Jaume IResumen
This project has been created by Adrián González Ramírez as his Final Degree Project of the Degree in Design and Development of Video Games of the Jaume I University [1].
This project focuses on creating a tower ... [+]
This project has been created by Adrián González Ramírez as his Final Degree Project of the Degree in Design and Development of Video Games of the Jaume I University [1].
This project focuses on creating a tower defense video game [2] and incorporating different
models graphs (Neural Networks [3]) generated with Machine Learning Agents [4] and the Proximal Policy Optimization (PPO) [5] Reinforcement Learning [6] algorithm, which will determine
the behavior of the Non Playable Characters (NPC).
Machine Learning [7] is, in the decade of 2010, a very present technique in a large number of
areas. The basis of the machine learning studied in this project are the same for any project that
uses Reinforcement Learning [6].
The method presented in this project shows a way to get different difficulty levels without
hardcoded behaviors, as Rubber banding [8], making less evident the manipulation of the difficulty
in the attempt to keep users desire to keep playing. [-]
Palabras clave / Materias
Descripción
Treball final de Grau en Disseny i Desenvolupament de Videojocs. Codi: VJ1241. Curs acadèmic: 2017/2018
Tipo de documento
info:eu-repo/semantics/bachelorThesisDerechos de acceso
info:eu-repo/semantics/openAccess
Aparece en las colecciones
El ítem tiene asociados los siguientes ficheros de licencia: