Authors : Dhaval Barai , Ashwini Hivse, Chetali Bhoyar, Pracheta Lokhande, Dr.Amol W. Pardhi
DOI : 10.46335/IJIES.2024.9.7.16
Article Link
Abstract - This paper proposes a novel approach for developing a dynamic AI opponent in a 2D fighting game using Stateful Learning and Unity Machine Learning Agents (ML-Agents). The AI leverages its observations of the player's recent attacks to adapt its defensive and offensive strategies, creating a more engaging and challenging opponent. This research details the design, implementation, and evaluation of this system within the Unity game engine, highlighting the chosen learning algorithm, data structures, and decision-making logic. The paper concludes by discussing potential future applications and areas for further research.
