https://arshadhs.github.io/tags/maximum-margin/Recent content in Maximum Margin on Arshad SiddiquiHugoen-usFri, 08 May 2026 00:00:00 +0000https://arshadhs.github.io/docs/ai/machine-learning/07-support-vector-machines/Fri, 08 May 2026 00:00:00 +0000https://arshadhs.github.io/docs/ai/machine-learning/07-support-vector-machines/<h1 id="support-vector-machine-svm"> Support Vector Machine (SVM) <a class="anchor" href="#support-vector-machine-svm">#</a> </h1> <p><strong>Support Vector Machine (SVM)</strong> is a <strong>supervised machine learning algorithm</strong> used for:</p> <ul> <li><strong>Classification</strong> (most common)</li> <li><strong>Regression</strong> (SVR – Support Vector Regression)</li> </ul> <p>It connects many earlier ideas:</p> <ul> <li>classification and decision boundaries</li> <li>linear classifiers</li> <li>margins</li> <li>optimisation</li> <li>constrained optimisation</li> <li>kernels for non-linear data</li> </ul> <p>SVM is a <strong>discriminative classifier</strong>.</p> <p>That means it does not try to model how each class is generated.</p> <p>Instead, it tries to find the <strong>best separating boundary</strong> between classes.</p>