SVM (Support Vector Machine)

A Support Vector Machine (SVM) is a supervised learning algorithm used for classification and regression tasks. It performs especially well on small, high-dimensional datasets. It is a flexible model that can handle both linear and non-linear classification, constructing a high-generalization classifier by maximizing the margin (distance) between the decision boundary (separating hyperplane) and the classes. Key characteristics of SVM: • Improves classification confidence by maximizing the margin between classes • The kernel trick allows data that cannot be linearly separated to be mapped to a higher-dimensional space for classification • A robust learning algorithm that is resistant to noise and outliers • Can also be applied to regression tasks (SVR: Support Vector Regression) Representative kernel functions include: • Linear Kernel • Polynomial Kernel • RBF Kernel (Radial Basis Function) • Sigmoid Kernel Use cases: • Handwritten character and image recognition (e.g., MNIST) • Email spam filtering • Binary classification problems where precision matters, such as medical diagnosis • Scenarios with many features but few samples, such as gene data classification SVM is a flagship algorithm from the pre-deep learning machine learning era and retains strong popularity today as a lightweight, high-accuracy classifier. Its ease of use with non-linear classification via the kernel method and theoretical stability keep it actively used across research, education, and industry.