Particle Swarm Optimization-based method for generating healthy lifestyle recommendations

Abstract

This paper addresses the problem of reducing the risk of developing cardiovascular diseases by proposing a nature-inspired method that generates healthy lifestyle recommendations. The method analyses the profile of a person including general health information, preferences and daily dietary and physical activities habits using a Naive Bayes classifier to identify unhealthy behaviours that increase the risk of developing cardiovascular diseases. If an unhealthy behaviour is detected, the method applies a Particle Swarm Optimization-based algorithm to generate healthy lifestyle recommendations that adjust the person's body mass index within healthy limits. The proposed method has been integrated into an experimental prototype and evaluated on a set of generated user profiles.