Reference no: EM133300676

TRUE/FALSE

1. For a rule-based classifier, if the rules are exhaustive, a test query will trigger exactly snerule.

2. For linearly separable problems, perceptron learning algorithm will create a linear hyperplane that separates the data perfectly.

3. Multilayer Artificial Neural Network can handle redundant and irrelevant attributes

4. Random forest algorithm uses bootstrap sample to test every decision tree

5. For imbalanced training set, to improve the classification performance, we can assign lower weights to instances from the majority class.