ml quiz 4

What is the primary purpose of Quadratic Discriminant Analysis (QDA)?

• To find the separating hyperplane in SVM
• To classify data points into two classes
• To generate non-linear decision boundaries
• To perform dimensionality reduction

In LDA, what happens when the mean of distributions is shared between classes?

• LDA can create a new axis for separation.
• LDA becomes computationally efficient.
• LDA fails to create a linearly separable axis.
• LDA produces a higher-dimensional output.

What is the primary difference between Linear Discriminant Analysis (LDA) and Flexible Discriminant Analysis (FDA)?

• LDA is used for binary classification, while FDA is for multi-class problems.
• LDA assumes linear groups of inputs, while FDA handles non-linear groups.
• LDA applies regularization, while FDA does not.
• LDA works only for image processing, while FDA is for finance.

ML estimation aims to maximize which function?

• Loss function
• Probability function
• Likelihood function
• Error function

What step follows the differentiation of the log-likelihood function with respect to θ in ML estimation?

• Constructing the likelihood function
• Taking the logarithm of the likelihood
• Setting the derivative to zero
• Using properties of i.i.d samples

In SVM, what is the primary goal of finding a hyperplane?

• Minimize the margin between classes
• Maximize the margin between classes
• Reduce the number of support vectors
• Create a non-linear decision boundary

What are the data points closest to the hyperplane called in SVM?

• Decision vectors
• Feature vectors
• Support vectors
• Kernel vectors

Which type of SVM is suitable for linearly separable data?

• Linear SVM
• Non-linear SVM
• Soft-margin SVM
• Hard-margin SVM

What is the main purpose of using a kernel in SVM?

• Reduce computational complexity
• Increase margin size
• Transform data into higher-dimensional space
• Simplify the hyperplane equation

Which type of SVM kernel is most useful for handling non-linear separation problems?

• Linear kernel
• Polynomial kernel
• Radial Basis Function (RBF) kernel
• Sigmoid kernel