Question

1) For a given text, the outcome of the sentiment analysis (unknown variable of sentiment) will be the same irrespective of the pre-built dictionaries used True or False

------------------------------------------------------------

------------------------------------------------------------

2) When growing a decision tree, when do you stop branching at a node?
a) The node is pure or the number of observations in the node is less than/equal to a preset threshold value

b) The node is impure or the number of observations in the node is less than/equal to a preset threshold value

c) The node is pure or the number of observations in the node is more than/equal to a preset threshold value

d) The node has a single observation

------------------------------------------------------------

------------------------------------------------------------

3) A small (close to 0) value of GINI index indicates that ______

a) The node is pure

b) The entropy of the node is very small

c) The node mostly contains observations from a single class

d) All of the above

------------------------------------------------------------

------------------------------------------------------------

4) Which of the following is true about Random Forests and Bagging?


1) In Random Forest, the decision trees are trained on a subset of the samples and a subset of the predictors at each split

2) In Random Forest, the decision trees are trained on a subset of the samples and the complete set of predictors at each split

3) In Bagging, the decision trees are trained on a subset of the samples and a subset of the predictors at each split

4) In Bagging, the decision trees are trained on a subset of the samples and the complete set of predictors at each split

a) 1 and 3

b) 2 and 3

c) 1 and 4

d) 2 and 4

Answer 1

1. False. The outcome will not be same and it will depend on the pre-built dictionaries.

2. The branching will stop when the node is pure or the number of observations in the node is less than/equal to the threshold value. A) is the correct option.

3. If gini index = 0, it means all the observations are from a single class or the node is pure hence small entropy. Hence, all the options are correct. D) is the correct option here.

4. In random forest, only a subset of features are used while in Bagging all the features are used. Hence, 1 and 4 are correct. Correct option is C)