Question

Please select all the statements that are true.

Question 3 options:

	For binary classification, both Logistic Regression and Support Vector Machine learn a linear boundary from the training data that try to separate the two classes.
	The linear decision boundary that is learned by Support Vector Machine from the training data maximizes the margin between each class to the boundary.
	Logistic Regression is more scale-able to big data than Support Vector Machine
	Logistic Regression has better generalization than Support Vector Machine

Answer 1

The linear decision boundary that is learned by Support Vector Machine from the training data maximizes the margin between each class to the boundary. True

ans- In SVM the main objective is to maximize the margin which is the distance between two separating hyperplane.

Logistic Regression is more scale-able to big data than Support Vector Machine True

ans- Since Logistic Regression models are simple compared to SVM, they are better suited when there is a huge amount of data. But if the data dimension are high compared to number of data sample then SVM would be a better model.

Logistic Regression has better generalization than Support Vector Machine False

ans- Since in SVM the margin maximizes they become less prone to overfitting, which is why when compared to logistic Regression SVMs have better generalisation i.e. they will perform better to unseen data.