1. : What is the primary application of Support Vector Machines (SVM) in data mining?
(A) Clustering
(B) Regression
(C) Classification and Regression
(D) Association rule mining
2. : What is the main objective of the SVM algorithm?
(A) To find the shortest distance between data points
(B) To find the optimal hyperplane that maximizes the margin between classes
(C) To cluster data points into different groups
(D) To reduce the dimensionality of the data
3. : In the context of SVM, what is a “support vector”?
(A) A data point that is closest to the hyperplane
(B) A data point that is furthest from the hyperplane
(C) A centroid of a cluster
(D) A decision node in a decision tree
4. : Which of the following is true about the kernel trick in SVM?
(A) It transforms the data into a lower-dimensional space
(B) It allows SVM to create non-linear decision boundaries
(C) It reduces the number of support vectors
(D) It normalizes the data
5. : What does the “C” parameter in SVM control?
(A) The width of the margin
(B) The type of kernel function
(C) The penalty for misclassified data points
(D) The number of support vectors
6. : Which of the following is NOT a commonly used kernel function in SVM?
(A) Linear kernel
(B) Polynomial kernel
(C) Gaussian (RBF) kernel
(D) Decision tree kernel
7. : What is the main advantage of using SVM for classification tasks?
(A) It is simple to implement
(B) It works well with high-dimensional data
(C) It requires a small amount of computational resources
(D) It does not need any tuning of parameters
8. : Which metric is maximized by the SVM algorithm to achieve the optimal hyperplane?
(A) Precision
(B) Recall
(C) Margin
(D) Support vector count
9. : How does the choice of the kernel function affect the performance of an SVM?
(A) It determines the computational complexity of the model
(B) It has no effect on the performance
(C) It influences the shape of the decision boundary
(D) It controls the regularization of the model
10. : What is the primary challenge when using SVM with very large datasets?
(A) Overfitting the training data
(B) High computational cost and memory usage
(C) Difficulty in choosing the right kernel
(D) Lack of interpretability of the model