Evolution of Artificial Intelligence Based Burned Calories Prediction System Using Novel Hybrid Learning Methodology

In this paper, we propose a novel approach for predicting burned calories through a hybrid learning methodology, termed NHLM, which integrates the power of AutoEncoder (AE) and EfficientNet. Accurate estimation of burned calories is crucial for individuals aiming for effective fitness management and weight loss. However, existing prediction models often struggle with precision and generalization. Our NHLM framework addresses these challenges by leveraging the strengths of both AE and EfficientNet. The AE component is employed for feature extraction and dimensionality reduction, enabling the model to capture essential characteristics from input data while minimizing information loss. Meanwhile, EfficientNet, renowned for its efficiency and accuracy in image classification tasks, acts as the primary predictive model within NHLM. The integration of these two components results in a synergistic effect, where the AE enhances the representation of input features, allowing EfficientNet to make more informed predictions. Furthermore, our NHLM architecture is trained on a diverse dataset consisting of various physical activities and corresponding calorie expenditure measurements, ensuring robustness and adaptability. Experimental results demonstrate the effectiveness of our proposed approach, achieving an impressive prediction accuracy of 97%. This high level of accuracy showcases the capability of NHLM in accurately estimating burned calories across different activities and individuals, thereby facilitating personalized fitness monitoring and optimization.