Framework for Predicting Large Datasets using Enhanced SVM-CRFE-GK Algorithm with Sparse Kernels

The advancement in predictive modeling for large datasets requires efficient methods that address dimensionality reduction, computational efficiency, and prediction accuracy. This paper proposes an Enhanced Support Vector Machine framework (SVM-CRFE-GK with Sparse Kernels) that integrates Chi-Recursive Feature Elimination (Chi-RFE) and Greedy Kernel optimization. By leveraging Sparse Kernels and utilizing cube products of support vectors, the framework efficiently handles large datasets while maintaining high accuracy and reduced computational overhead. The methodology was validated on a COVID-19 huge dataset, demonstrating superior performance in terms of accuracy, memory utilization, and processing time compared to conventional SVM methods. The proposed framework showcases its robustness in real-world applications, particularly in high-dimensional medical diagnostics.