vector machine (SVM)

资 源 简 介

In this paper， we show how support vector machine (SVM) can beemployed as a powerful tool for $k$-nearest neighbor (kNN)classifier. A novel multi-class dimensionality reduction approach，Discriminant Analysis via Support Vectors (SVDA)， is introduced byusi

详 情 说 明

In this paper, we present a detailed analysis on how support vector machine (SVM) can be utilized as a powerful tool for the $k$-nearest neighbor (kNN) classifier. Our approach involves introducing a novel multi-class dimensionality reduction technique, called Discriminant Analysis via Support Vectors (SVDA), which makes use of SVM. This technique is particularly useful in deriving the non-linear version, Kernel Discriminant via Support Vectors (SVKD), through kernel mapping.

One of the key advantages of SVDA is that it only involves support vectors to obtain the transformation matrix, thereby significantly reducing computational complexity for kernel-based feature extraction. Our experiments, which were conducted on several standard databases, clearly demonstrate a significant improvement in LDA-based recognition. Therefore, our proposed SVDA approach offers a promising solution for improving the efficiency and accuracy of kNN classification.

Moreover, we also discuss the potential applications of SVM in various other fields, such as image recognition, natural language processing, and speech recognition. We highlight the key benefits of using SVM, such as its ability to handle high-dimensional data, its robustness against noise and outliers, and its flexibility in dealing with non-linearly separable data. Overall, our findings suggest that SVM is a highly versatile and effective tool for a wide range of classification problems in different domains.