Please use this identifier to cite or link to this item: https://www.um.edu.mt/library/oar/handle/123456789/74758
Title: Conflict of nations
Authors: Kudde, Dale (2019)
Keywords: Computer games
Artificial intelligence
Machine learning
Issue Date: 2019
Citation: Kudde, D. (2019). Conflict of nations (Bachelor's dissertation).
Abstract: One of the most prevailing problems the gaming industry faces today is designing A.I systems that produce non-player characters (NPCs) which are capable of exhibiting intelligent and adaptive behaviour in ever more complex environments. This is especially true when it comes to real-time strategy games (RTS) were the game environment tend to be more complex and generally boast a larger action-state space than most other game genres. Whilst such environments make for a complex, well detailed games with in-depth gameplay such as StarCraft, Age Of Empires and Tropico, they also impose great restrictions when it comes to designing A.I systems within them. It is important to note that the primary goal of any game is first and foremost to provide entertainment to the end user as such any A.I system must not only be capable of adapting to ever changing circumstances but also sufficiently provide enough of a challenge to the players to keep them alerted and engaged while not challenging enough to completely overwhelm them. To this end we introduce CONAI, an artifi cial system set in the online RTS game called Conflict Of Nations. CONAI is an AI system working within a goal-driven autonomy (GDA) framework which uses reinforcement learning to learn the most appropriate strategies at any given situation.
Description: B.SC.ICT(HONS)ARTIFICIAL INTELLIGENCE
URI: https://www.um.edu.mt/library/oar/handle/123456789/74758
Appears in Collections:Dissertations - FacICT - 2019
Dissertations - FacICTAI - 2019

Files in This Item:
File Description SizeFormat 
Kudde Dale.pdf
  Restricted Access
1.4 MBAdobe PDFView/Open Request a copy


Items in OAR@UM are protected by copyright, with all rights reserved, unless otherwise indicated.