machine-learning SVM Difference between logistic regression and SVM


Example

Decision boundary when we classify using logistic regression- Logistic Regression

Decision boundary when we classify using SVM-

Classification using SVM

As it can be observed, SVM tries to maintain a 'gap' on either side of the decision boundary. This proves helpful when we encounter new data.

With new data-

Logistic regression performs poorly (new red circle is classified as blue) -

New data (red circle) with logistic regression's decision boundary

Whereas SVM can classify it correctly (the new red circle is classified correctly in red side)-

New red circle is classified correctly in SVM