Scribd
Upload
9 views3 pages

Week 6

Uploaded by

Raja mitme
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
9 views3 pages

Week 6

Uploaded by

Raja mitme
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Introduction to Machine Learning

Week 6
Prof. B. Ravindran, IIT Madras

1. (1 Mark) Entropy for a 90-10 split between two classes is:


(a) 0.469
(b) 0.195
(c) 0.204
(d) None of the above
Sol. (a)

2. (2 Mark) Consider a dataset with only one attribute(categorical). Suppose, there are 8 un-
ordered values in this attribute, how many possible combinations are needed to find the best
split-point for building the decision tree classifier?
(a) 511
(b) 1023
(c) 512
(d) 127
Sol. (d)
Suppose we have q unordered values; the total possible splits would be 2q−1 − 1. Thus, in our
case, it will be 27 − 1 = 127.
3. (2 mark) Having built a decision tree, we are using reduced error pruning to reduce the size of
the tree. We select a node to collapse. For this particular node, on the left branch, there are
three training data points with the following outputs: 5, 7, 9.6, and for the right branch, there
are four training data points with the following outputs: 8.7, 9.8, 10.5, 11. The average value
of the outputs of data points denotes the response of a branch. The original responses for data
points along the two branches (left & right respectively) were response left and, response right
and the new response after collapsing the node is response new. What are the values for
response left, response right and response new (numbers in the option are given in the same
order)?
(a) 9.6, 11, 10.4
(b) 7.2; 10; 8.8
(c) 5, 10.5, 15
(d) Depends on the tree height.
Sol. (b)

4. (1 Mark) Which of the following is a good strategy for reducing the variance in a decision tree?

1
(a) If improvement of taking any split is very small, don’t make a split. (Early Stopping)
(b) Stop splitting a leaf when the number of points is less than a set threshold K.
(c) Stop splitting all leaves in the decision tree when any one leaf has less than a set threshold
K points.
(d) None of the Above.

Sol. (b)

5. (1 Mark) Which of the following statements about multiway splits in decision trees with cate-
gorical features is correct?
(a) They always result in deeper trees compared to binary splits
(b) They always provide better interpretability than binary splits
(c) They can lead to overfitting when dealing with high-cardinality categorical features
(d) They are computationally less expensive than binary splits for all categorical features
Sol. (c)

6. (1 Mark) Which of the following statements about imputation in data preprocessing is most
accurate?
(a) Mean imputation is always the best method for handling missing numerical data
(b) Imputation should always be performed after splitting the data into training and test sets
(c) Missing data is best handled by simply removing all rows with any missing values
(d) Multiple imputation typically produces less biased estimates than single imputation meth-
ods
Sol. (d)

7. (2 Marks) Consider the following dataset:

feature1 feature2 output


18.3 187.6 a
14.7 184.9 a
19.4 193.3 a
17.9 180.5 a
19.1 189.1 a
17.6 191.9 b
19.9 190.2 b
17.3 198.6 b
18.7 182.6 b
15.2 187.3 b

Which among the following split-points for feature2 would give the best split according to the
misclassification error?
(a) 186.5

2
(b) 188.6
(c) 189.2
(d) 198.1
Sol. (c)

You might also like