Support Vector Machine

Question 1

Support Vector Machine

Answer 1

To establish the broadest possible “street” between distinct classes.

Question 2

Support vector

Answer 2

Any instance located on the “street”.

Question 3

Decision boundary

Answer 3

Entirely determined by the support vectors

Question 4

2 Type of Support Vector Machine

Answer 4

SVM Classification - Fit the largest possible street between two classes while limiting margin violation
SVM Regression - Fit as many instances as possible on the street while limiting margin violations

Question 5

2 Type of Support Vector Machine Margins (Only for classification)

Answer 5

Hard Margin Classification
Soft Margin Classification - To find a balance between maintaining the street as wide as possible while also limiting the number of margin violations.

Question 6

2 Characteristic of Hard Margin Classification

Answer 6

All instances must be off the street
All instances must be on the right side

Question 7

2 Issue of Hard Margin Classification

Answer 7

Works only when data is linearly separable.
Sensitive to outliers.

Question 8

2 Approach of Nonlinear SVM Classification

Answer 8

Approach 1 - High polynomial degree
Approach 2 - Polynomial Kernel (kernel trick) - Higher dimensional; Not add new features

Question 9

C (Regularization Hyperparameter)

Answer 9

Controls the trade-off between achieving a large margin and minimising classification errors

Small value of C - Wider margin & some misclassifications (margin violations)
Large value of C - Narrower margin & potentially overfit