Multi-Class SVM Classifier MATLAB Implementation

MATLAB code for multi-class Support Vector Machine classifier implementation using machine learning algorithms for accurate classification across multiple categories

Support Vector Machine (SVM) for Enhanced k-Nearest Neighbor Classification with Dimensionality Reduction

This paper demonstrates the application of Support Vector Machine (SVM) as a robust foundation for improving k-nearest neighbor (kNN) classifiers. We introduce Discriminant Analysis via Support Vectors (SVDA), a novel multi-class dimensionality reduction technique that leverages SVM principles. The implementation involves using only support vectors to compute transformation matrices, reducing computational overhead for kernel-based feature extraction. Our methodology extends to non-linear versions through kernel mapping, achieving improved recognition performance in experimental validations across standard datasets.