Food Classification Using a Hybrid Framework with Transfer Learning and Machine Learning Models

This paper presents a hybrid framework for food classification that integrates transfer learning and machine learning techniques. To establish a baseline, a conventional CNN was implemented and evaluated as a base model for comparison. The framework used three transfer learning modelsEfficientNet, DenseNet, and MobileNet-for feature extraction, leveraging their capabilities to handle complex and highdimensional food image datasets. The extracted features are classified using machine learning models, including XGBoost and Random Forest, to achieve robust and scalable predictions. By combining the feature extraction strengths of deep learning with the interpretability and efficiency of machine learning classifiers, the proposed approach addresses the limitations of traditional methods. Experimental results demonstrate the framework’s superiority in terms of classification accuracy, precision, and robustness against noisy and low-quality data. These findings underscore the framework’s potential for real-time applications such as food safety monitoring and nutritional assessment.