An AI-Driven Personalized Exercise Recommendation System Using Clustering and Machine Learning Approach

Abstract

Regular physical activity is crucial for maintaining health and preventing chronic diseases. However, individuals often struggle to adhere to exercise routines due to lack of motivation, time constraints, and difficulty in finding appropriate activities. This study presents the development of a personalized Exercise Habit App that aims to promote regular physical activity by providing tailored exercise recommendations and support. The app leverages user profiling and clustering techniques to group users based on their demographic information, health conditions, and exercise preferences. Through a user-friendly interface developed using Flutter and the Dart programming language, the app guides users in setting exercise goals, selecting preferred activities, and tracking their progress. Personalized exercise recommendations are generated based on the user's cluster, taking into account their specific needs and preferences, as well as physical activity guidelines from authoritative organizations like the World Health Organization (WHO), American College of Sports Medicine (ACSM), and U.S. Department of Health and Human Services. The app aims to provide exercise recommendations that are tailored not only to individual preferences but also to specific health conditions and disease states, promoting safe and effective physical activity for users with chronic conditions. Additionally, the app will incorporate collaborative filtering and association rule mining techniques to further enhance the recommendation system by leveraging user-item interactions and identifying patterns in exercise preferences. The Exercise Habit App, implemented using Flutter for cross-platform compatibility and Hive for efficient data management, has the potential to encourage and support individuals in establishing and maintaining an exercise routine, ultimately leading to improved overall health and well-being.