In 2008, Maurice Bergsma and Pieter Spronck proposed an AI architecture, named ADAPTA, for a deeply simplified version of the Turn Based Strategy (TBS) videogame Advance WarsTM. The aim of this thesis is to extend the concepts of such architecture and apply them to the original game. The created AI is able to change its behavior and strategies between matches: its aim is not necessarily to win, but is parametrized to give the player a customized challenge, in line with the Game Design principle of Flow. Since the ADAPTA architecture features modularity as one of its strongest points, this work also attempts to lay the foundations for future research. As an example of this, a new type of approach in this field is also proposed, which makes use of a Convolutional Neural Network applied to a TBS game.
Nel 2008, Maurice Bergsma e Pieter Spronck proposero un’architettura, nominata ADAPTA, per un Intelligenza Artificiale applicata ad una versione estremamente semplificata del videogioco strategico a turni Advance WarsTM. Lo scopo di questa tesi è di estendere i concetti di tale architettura e applicarli al gioco originale. La IA creata può modificare il suo comportamento e le sue strategie tra una partita e l’altra: lo scopo non è necessariamente vincere, ma è parametrizzato per poter fornire al giocatore un’esperienza personalizzata, in linea con il principio di Flow nell’ambito di Game Design. Dato che l’architettura ADAPTA è caratterizzata da una forte modularità, questa tesi vuole anche proporsi come possibile fondamento per ricerche future. A dimostrazione di questo viene anche proposto un nuovo tipo di approccio in questo campo, che utilizza Convolutional Neural Networks applicate ad un gioco strategico a turni.
Extension of the ADAPTA architecture applied to the videogame Advance Wars
Carnaghi, Lorenzo
2020/2021
Abstract
In 2008, Maurice Bergsma and Pieter Spronck proposed an AI architecture, named ADAPTA, for a deeply simplified version of the Turn Based Strategy (TBS) videogame Advance WarsTM. The aim of this thesis is to extend the concepts of such architecture and apply them to the original game. The created AI is able to change its behavior and strategies between matches: its aim is not necessarily to win, but is parametrized to give the player a customized challenge, in line with the Game Design principle of Flow. Since the ADAPTA architecture features modularity as one of its strongest points, this work also attempts to lay the foundations for future research. As an example of this, a new type of approach in this field is also proposed, which makes use of a Convolutional Neural Network applied to a TBS game.File | Dimensione | Formato | |
---|---|---|---|
2022_04_Carnaghi.pdf
accessibile in internet per tutti
Descrizione: Tesi Carnaghi Lorenzo
Dimensione
3.78 MB
Formato
Adobe PDF
|
3.78 MB | Adobe PDF | Visualizza/Apri |
executive_summary_Carnaghi_Lorenzo.pdf
accessibile in internet per tutti
Descrizione: Executive Summary Carnaghi Lorenzo
Dimensione
1.28 MB
Formato
Adobe PDF
|
1.28 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/187184