Abstract:
This research seeks to create and operationalize a system labeled AI-powered Fitness Guidance and Analytics System that will enable gym and fitness center owners to im prove operational efficiency increase engagement among clients and enable enhanced decision-making. Leveraging sophisticated machine learning algorithms known as XG Boost and other complementory predictive techniques the system seeks to estimate sales on a monthly basis detect members who may discontinu servic and profile ones behavior relative to their attendance, demographics, and subscription history. A consolidated web interface constructed using Next.js and Tailwind CSS supplies users with operational dashboards that offer predictive analytics, visual dashboards, and alerts on members at risk of disengagement. These functionalities plug into the overall objective of supporting users in their pursuit of evidence-based interventions. This is achieved within the systems architecture through providing interpretability of the models in a detailed manner such as through visualizing features that matter most as well as providing users with predictive analytics at a models output. The system recorded 85% accuracy in sales prediction and 0.88 AUC score in predicting members that were at risk of disengaging. It is indeed the user experience fused with machine learning and analytics that demonstrates evidence-based real-time decision making and frictionless interface that the analytics are designed to ease in the first place the power of AI in the fitness analytics space
