Support Vector Machine

lastmod 2026-08-26created 2014-10-2527 edits

A support vector machine (SVM) is a method used in supervised learning, and a basic yet very well-known Classification model. Among the hyperplanes that separate the given data, it finds the hyperplane (Optimal hyperplane) farthest from the data.

This technique can also be applied to nonlinear classification by using the kernel trick.

Overview

• SVM Algorithm
• (ML) Deep learning model - SVM

SVM (Support Vector Machine) is an algorithm that finds the boundary line (plane) that best classifies distributions of messy patterns, systematized so that it can be extended to multidimensional vectors (patterns). In more complex settings, it is even used for speech recognition. First, look at the figure below for an intuitive understanding. You can see several points distributed in a 2D plane, and there are several lines that can classify these points (patterns). The key question is how to find the lines (called Support vectors) that serve as the classification boundary in a multidimensional space. SVM is the algorithm that provides the answer. In a sense, whereas a decision tree finds classification criteria based on levels or depth, SVM finds classification criteria in the multidimensional space where the patterns are distributed, so the two can be seen as similar.

SVM terms

• Hyperplane (Optimal hyperplane)
• Kernel function
• Training (Train)
• Prediction (Predict)

Real-time image-based automatic detection method, apparatus, and system for flame, smoke, and object movement

• Korean title: Real-time image-based automatic detection method, apparatus, and system for flame, smoke, and object movement
• English title: AUTOMATIC DETECTION METHOD, APPARATUS AND SYSTEM OF FLAME, SMOKE AND OBJECT MOVEMENT BASED ON REAL TIME IMAGES
• Registration number (date): 1015811620000 (2015.12.22)
  • Patent application No. 10-1581162

See also

• libsvm
• Neural networks: one disadvantage of neural networks is that they stop learning once they have learned all the training data.
• Radial basis function (RBF)
• OpenCV:SVM
• Classification model
• KNN

Favorite site

• Wikipedia (en) SVM
• Defending against drive-by-download using SVM (support vector machine)
• SVM (Support Vector Machine)
• [Recommended] Machine learning - Supervised learning - Support Vector Machine (SVM)
• [Recommended] Machine learning by implementing SVM (support vector machine) with dlib (data mining)
• Implementing a Support Vector Machine (SVM)
• 2-class classifier using libsvm (1) - What is SVM
• Computer Vision and Pattern Recognition Society - Support Vector Machines ¹
• SVM glossary
• Kernel functions and support vector machines (SVM)
• Introduction to the SVM machine learning algorithm
• Programming Collective Intelligence study: 9. Advanced classification: kernel methods and SVM
• [Recommended] cheezestick story - Deriving the SVM (Support Vector Machine) hyperplane ²
• Introduction to machine learning models (2) - Support Vector Machine (SVM)

Machine learning - Supervised learning

• Introduction to BCI-Machine learning-Supervised learning-1 overview ³
• Introduction to BCI-Machine learning-Supervised learning-2 Artificial neural network (ANN) ⁴
• Introduction to BCI-Machine learning-Supervised learning-3 Support Vector Machine (SVM) ⁵

Software

• SVM - Support Vector Machines Software
• SVMLight
• LIBSVM
• mySVM
• BioSVM

References