Identify the following as True or False: (1) Naive Bayes is a linear classifier. (2) SVMs are only usable when the classes are linearly separable in the feature space. (3) Adding training data always results in a monotonic increase in the accuracy of a Naive Bayes classifier. (4) When sufficient data is available, SVMs generally perform as well or better than other common classifiers such as KNN. (5) With enough training data, the error of a nearest neighbor classifier always goes down to zero.

1. True: Naive Bayes is a linear classifier.

2. False: SVMs are only usable when the classes are linearly separable in the feature space.

3. False: Adding training data always results in a monotonic increase in the accuracy of a Naive Bayes classifier.

4. True: When sufficient data is available, SVMs generally perform as well or better than other common classifiers such as KNN.

5. False: With enough training data, the error of a nearest neighbor classifier always goes down to zero.

Explanation:

Naive Bayes is a linear classifier that leads to a linear decision boundary. It can be applied to a linearly separable problems and when the elements are independent i. e the occurrence of an element doesn't affect the occurrence of another. It can be used for making multi class predictions in artificial intelligence.

The Support Vector Machine (SVM) on the other hand, can either be a non-linear classifier (with RBF kernel) or a linear classifier (with linear kernel). It maximizes the margin of a decision boundary in its mode of operation.

Hence, the SVMs can be used for regression or classification problems.

For example, determining whether an e-mail is a spam or not.