International Journal of Scientific Research and Engineering Development

At present, people lead a busy schedule because of changes in lifestyle and work patterns. Although exercising helps a lot in maintaining a healthy physique and sound psychological conditions, people are not largely abiding by this regime. Along with this, unsound eating habits and a sedentary lifestyle are leading to an immense rise in obesity cases, making this a grave concern in today’s community. Therefore, people are now largely dealing with obesity or weight-related problems in an attempt to regulate their eating habits and exercising regime. Although calculating calorie consumption isn’t a tough task since calorie information is easily accessible on food items and websites, calculating calorie burning during exercising regimes is a tough task and not always accurate [2]. The prime aim of this research work is to compare machine learning algorithms, as well as design an accurate predictive model to estimate calories burned during physical activity. The designed system uses parameters like heart rates, body tempera- ture, weight, height, age, gender, and exercise duration to estimate energy consumption [1]. Among all parameters used, heart rates are one of the most crucial parameters, as they indicate actual levels of physical activity. The traditional approach to estimate calories uses predefined mathematical formulas that have an average value without considering levels of physical activity or individual variations. Thus, it may lead to incorrect outputs [5]. To counter these issues, supervised machine learning al- gorithms are developed and tested using the dataset that is correlated to exercises. The generated algorithms are then measured for their efficiency using regression parameters like Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE) [7]. From the results obtained by carrying out experiments, it has been proved that machine learning algorithms perform better than traditional approaches for accurate predictions regarding calorie calculation with greater personalization. These can then be applied to healthrelated applications for a healthy life.

Dr.T.Subba Reddy, G.Priyanka, Kajal Kumari, A.Naga Sri Lekha, D.Hepsiba,"Caloric Expenditure Estimation From Exercise Duration and Physiological Metrics Using XGBoost Regression" International Journal of Scientific Research and Engineering Development, V9(2): Page(960-964) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.