Question

What is the difference and similarities between SVM and Logistic Regression?

Answer 1

Logistic regression is an algorithm used to solve classification problems. It helps to predict the future value and explains relationship between variables. Here, the output variable is in binary form that is ranges between 0 and 1.

SVM or support vector machine is used for both classification and regression analysis. It pays more attention to classification problems. Here, the data is seperated into classes by using hyperplane or decision boundary. This makes complex non linear functions more easy to analyse.

The similarities between SVM and Logistic regression are-

1. Logistic regression and SVM both are uesd to make linear decision boundary.

2. Logistic regression are trained to maximise the likelihood and SVM are trained to find the maximum margin that seperates the classes.

The differences between SVM and logistic regression are-

SVM	Logistic Regression
1. It helps to find the best margins that seperates the classes so that the risk of error is reduced.	1.It has different margins with different weights near optimal point.
2. Unstructured and semi-structured data are used here.	2. Independent variables are used here.
3. The risk of overfitting is less.	3. The risk of overfitting is more.