Intelligent Multi-Agent Reinforcement Learning Based Disease Prediction and Treatment Recommendation Model
Thota Radha Rajesh, Surendran Rajendran
Abstract
The problem of disease prediction and recommendation has been well studied. There exist several techniques for disease prediction but suffer to achieve the expected performance. Towards the problem, an efficient Intelligent Multi-Agent Reinforcement Learning based Disease Prediction and Treatment Recommendation (IMRLDPTR) is presented. The model maintains an agent container with several mobile agents and fetches data from different locations of the network and cloud. Using the data fetched, the method generates a rule set for various disease classes. During the test phase, the method reads the input samples and symptoms to compute Disease Attraction weight (DAW) towards various classes. Based on the value of DAW, the method performs disease prediction and measures the Disease Curing Rate (DCR) for different classes of treatment. Based on the value of DCR, a set of recommendations are generated to support the problem.