Employing Machine Learning Techniques to Categorize users in a Fitness Application

Abstract

Nowadays, health is a top priority. People are putting in a lot of effort to improve their health and make their bodies healthier. The majority of people use fitness apps to track their daily activities. This Fitness App provides all users with one-stop exercise solutions such as fitness training, cycling, running, yoga, and fitness diet guidance. The Fitbit Kaggle dataset, which contains 18 CSV files and approximately 2.5K users, was used in this study. The data set was analyzed in terms of “sleep vs active minutes” and “logged activity vs not logged activity.” The K-means machine learning technique is used to cluster App users based on a variety of factors, and whether they are eligible for bonuses or reward points. This paper's research focused on user categorization using unsupervised learning based on cluster. Such a Fitness App integrated with machine learning technique could intelligently motivated their customer in staying active throughout the day.