Question

In each of the classification scenarios listed below, you are given a set of classifiers and a description of the classification scenario. For each scenario, state the choice of the classifier that is best suited for the dataset along with a brief explanation supporting your answer.

a) Scenario: Data contains some missing values for certain attributes in the training and test data. Classifiers available: Naïve Bayes, RIPPER, KNN

b) Scenario: Many of the attributes are irrelevant (contain no information about the class). Classifiers available: KNN, Decision Trees

c) Scenario: Dataset contains attributes that are not discriminative by themselves but are discriminative in combination. Classifiers available: KNN, Naïve Bayes, Decision Trees (taking

single attribute at a time)

Answer 1

a. For this case, we can use the Naive Bayes Classification algorithm, because it is the one in which it does not depend on all the attributes to contribute to the final probability. All the features independently contribute to the final probability.

b. In this case, we should use "Decision Tree" because with this the root node will be made by the attribute which is most important and decision tree will keep on eliminating the irrelevant attributes by making them at the other part of the tree. Thus Decision Tree is the correct classifier for this case.

c. For this case, KNN would be the best-suited algorithm as in this the labels which are present close to the one which we are trying to find will work together. Thus KNN would be the best-suited algorithm for this case.

Thanks