Effective Patient Scheduling For Public Health Events Based On Deep Reinforcement Learning

Abstract

Amidst the current worldwide pandemic of the novel coronavirus, there are still issues about the efficient allocation of emergency supplies and shortcomings in recovery protocols. To enhance individuals' health and overall welfare, it is imperative to enhance and refine the emergency patient scheduling framework for health emergencies while upholding the idea of a society with an inherent future for humanity. The paper proposed an Efficient Patient Scheduling for Public Health Emergencies (EPS-PHE) using Multiagent Reinforcement Learning (MRL). Originally intended to solve the PHE and improve EPS, the technology was conceived as a distributed multiagent system. MRL has proposed a decentralized incentive system evaluation technique for the EPS-PHE input. An incentive type called a Direct Incentive (DI) is used to show the relationships between individuals or details about two closely connected events. The quality of data transmission in EPS-PHE or the efficacy of the emergency network is measured using a Collective Incentive (CI). Simulation research has shown the value of the Deep Learning (DL) - MRL approach in EPS-PHE framework optimization.