Recommender System with Artificial Intelligence for Fitness Assistance System

This paper proposes a recommender system (RS) to support the fitness assistance system (FAS) with artificial intelligence. The RS is applied to make these suggestions for the beginners and existing users. The goal of the paper aims to develop an RS that has an ability to learn, analyze, predict, and make these suggestions as well as communicate to human through AI. Artificial Neural Network and Logistic Regression have been employed to predict the suitable workout for each beginner. In addition, the agent developed with reinforcement learning capability of Soar architecture help the members select their workout based on their condition. Through the experimental result, the effectiveness of utility application is validated.