Implementing Artificial Intelligence Agent Within Connect 4 Using Unity3d and Machine Learning Concepts
Nirmal Baby1, Bhargavi Goswami2

1Nirmal Baby, Department of Computer Science, Christ Deemed To Be University, Bangalore (Karnataka), India.
2Bhargavi Goswami, Department of Computer Science, Christ Deemed To Be University, Bangalore (Karnataka), India.
Manuscript received on 03 May 2019 | Revised Manuscript received on 15 May 2019 | Manuscript Published on 23 May 2019 | PP: 193-200 | Volume-7 Issue-6S5 April 2019 | Retrieval Number: F10320476S519/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Nowadays, we come across games that have unbelievably realistic graphics that it usually becomes hard to distinguish between reality and the virtual world when we are exposed to a virtual reality gaming console. Implementing the concepts of Artificial Intelligence (AI) and Machine-Learning (ML) makes the game self-sustainable and way too intelligent on its own, by making use of self-learning methodologies which can give the user a better gaming experience. The use of AI and ML in games can give a better dimension to the gaming experience in general as the virtual world can behave unpredictably, thus improving the overall stigma of the game. In this paper, we have implemented ‘Connect-4’, a multiplayer game, using ML concepts in Unity3D. The machine learning toolkit ‘ML-Agents’, which depends on Reinforcement Learning (RL) technique, is provided using Unity3D. This toolkit is used for training the game agent which can distinguish its good moves and mistakes while training, so that the agent will not go for same mistakes over and over during actual game with human player. With this paper, authors have increased intelligence of game agent of Connect 4 using Reinforcement Learning, Unity3D and ML-Agents toolkit.
Keywords: Artificial Intelligence, Machine Learning, Connect four, Game theory, Reinforcement Learning, Unity3D, ML-Agents.
Scope of the Article: Artificial Intelligence