18 Questions
What happens to the margin position of the hyperplane if the support vectors are deleted?
The margin position remains unchanged
In a linearly separable dataset, why are not all points on the supporting hyperplanes considered support vectors?
Because they do not influence the classifier's decision boundary
What is the main goal when searching for the largest margin classifier in SVM?
To maximize the distance between the classes
How are the weights in SVM determined during optimization?
Only the support vectors determine the weights
What does the hyperplane H0 represent in SVM?
The median plane between H1 and H2
What is the key difference between Maximum Margin Classifiers and Support Vector Classifiers?
Allowing misclassification
Why are support vectors crucial in determining the decision boundary in SVM?
They define the boundaries of each class
Why are Maximum Margin Classifiers very sensitive to outliers?
They optimize the margin based on all data points
What is the purpose of allowing some misclassification in Support Vector Classifiers?
To improve the generalization ability
How does a hard margin differ from a soft margin in Support Vector Machines?
Soft margin allows some errors for better generalization
What trade-off is involved when using a soft margin in Support Vector Machines?
Reduced sensitivity to outliers at the expense of model complexity
How is the best soft margin determined in Support Vector Machines?
Through cross-validation to find the best trade-off between errors and margin size
What is the formula for the margin of a separating hyperplane in Support Vector Machines?
(d+) + (d-)
Why should the decision boundary in SVM be as far away from the data of both classes as possible?
To maximize the margin
In SVM, how can a data point be identified as a support vector?
By removing it resulting in increasing the margin
What is the equation that corresponds to the decision boundary in SVM when trained on a dataset with 6 points, containing three samples with class label -1 and three samples with class label +1?
𝑊𝑋 + b = 0
What is the characteristic of the 𝑊 vector in SVM?
Orthogonal to H1 line
Why are methods like Lagrange multiplier method used in practice to solve SVM?
To handle non-linearly separable data
Explore the concept of Maximum Margin Classifiers in Support Vector Machines and how outliers can impact classification. Learn about the importance of maximizing margin for accurate classification.
Make Your Own Quizzes and Flashcards
Convert your notes into interactive study material.
Get started for free
