Reinforcement Learning in Supermarket Tycoon to Enhance Multitasking and Problem-Solving in Autistic Children

This research investigates the application of Multi-Agent Reinforcement Learning in the context of game therapy for children with autism. Specifically, it focuses on the utilisation of a mini supermarket management game as a therapeutic tool. This study utilise an actor-critic algorithm to conduct a comparative analysis of Independent Q-Learning (IQL) and multiagent actor-critic (MAAC) approaches. The objective of the game is to improve cognitive abilities and the coordination between visual perception and motor skills. The outcomes encompass enhanced financial management, organisational proficiencies, multitasking abilities, problem-solving aptitude, and observational acumen. The evaluation demonstrates that MAAC has a higher level of performance compared to other entities. The algorithm exhibits potential in facilitating the enhancement of skills and cognitive abilities in children with autism through the utilisation of captivating gameplay. The study makes a valuable contribution to the field of reinforcement learning by emphasising the effectiveness of Multi-Agent Actor-Critic (MAAC) algorithms. Additionally, it highlights the possibility of game therapy as a means of supporting cognitive development in individuals with autism. Subsequent investigations may focus on the enhancement of algorithms and the exploration of wider therapeutic uses.