A Machine Learning-Based Approach for Personalized Calorie Expenditure Prediction with an Integrated AI Fitness Chatbot

In an era where sedentary lifestyles are increasingly prevalent, the need for effective tools to manage personal fitness has never been more critical. This paper presents the design and implementation of a "Personal Fitness Chart," a web-based application designed to predict calorie expenditure with a high degree of personalization. The system leverages a Gradient Boosting Regressor model, a powerful machine learning algorithm, trained on a comprehensive dataset encompassing user attributes such as age, gender, height, weight, and exercise metrics including duration, heart rate, and body temperature. The web application, developed using the Streamlit framework, offers an intuitive user interface for data input and provides real-time predictions of calorie burn. A key innovation of this system is its integrated AI chatbot, powered by large language models via the OpenRouter API, which delivers personalized fitness recommendations based on user data and prediction results. Furthermore, the application can generate and email a personalized PDF fitness summary complete with user data, prediction results, a visual chart, and the full AI chatbot conversation offering users a tangible and comprehensive record of their session. This research demonstrates the significant potential of combining predictive machine learning with generative AI to provide tailored fitness guidance, thereby empowering individuals to take a more active role in their health and well-being.