ANALYSIS OF THE EFFECTIVENESS OF SUPPORT VECTOR MACHINES IN DATA CLASSIFICATION AND THEIR PRACTICAL APPLICATIONS

Hejjo Muhsen

The aim of the study is to improve the efficiency of unmanned aerial vehicles (One of the classification methods presented in this work is the support vector machine classifier. The support vector machine (SVM) is a supervised learning method that ensures a correspondence between desired input and output data. This method is based on statistical learning theory and is applied to classification tasks such as disease diagnosis, image classification, and handwritten text recognition. Traditional artificial neural networks face challenges in generalization because they rely on the principle of empirical risk minimization (ERM). Consequently, in 1995, Vapnik developed the support vector machine method to enhance the generalization process based on the principle of structural risk minimization (SRM). This principle surpasses empirical risk minimization, as it focuses on reducing the upper bound of expected risk rather than merely minimizing errors in the training set. This distinction provides the support vector machine with greater generalization capabilities, which is a key goal of statistical learning.

Keywords:Support vectors, support vector machine, Lagrange multiplier, kernel trick, maximum margin