Scribd
Upload
383 views2 pages

Week 2262666362hs

Uploaded by

bacoj54193
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
383 views2 pages

Week 2262666362hs

Uploaded by

bacoj54193
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/ 2

Introduction to Machine Learning

Week 2
Prof. B. Ravindran, IIT Madras

1. (1 Mark) State True or False: Typically, linear regression tend to underperform compared
to k-nearest neighbor algorithms when dealing with high-dimensional input spaces.
(a) True
(b) False
Sol. b
2. (2 Marks) Given the following dataset, find the uni-variate regression function that best fits
the dataset.

X Y
2 5.5
3 6.5
4 9
10 18.5

(a) f (x) = 1 × x + 4
(b) f (x) = 1 × x + 5
(c) f (x) = 1.5 × x + 3
(d) f (x) = 2 × x + 1

Soln. C
3. (1 Mark) Given a training data set of 500 instances, with each input instance having 6 di-
mensions and each output being a scalar value, the dimensions of the design matrix used in
applying linear regression to this data is

(a) 500 × 6
(b) 500 × 7
(c) 500 × 62
(d) None of the above

Soln. B
4. (1 Mark) Assertion A: Binary encoding is usually preferred over One-hot encoding to repre-
sent categorical data (eg. colors, gender etc)
Reason R: Binary encoding is more memory efficient when compared to One-hot encoding

(a) Both A and R are true and R is the correct explanation of A


(b) Both A and R are true but R is not the correct explanation of A
(c) A is true but R is false

1
(d) A is false but R is true

Soln. D
5. (1 Mark) Select the TRUE statement.
(a) Subset selection methods are more likely to improve test error by only focussing on the
most important features and by reducing variance in the fit.
(b) Subset selection methods are more likely to improve train error by only focussing on the
most important features and by reducing variance in the fit.
(c) Subset selection methods are more likely to improve both test and train error by focussing
on the most important features and by reducing variance in the fit.
(d) Subset selection methods don’t help in performance gain in any way.
Sol. a
6. (1 Mark) Rank the 3 subset selection methods in terms of computational efficiency:
(a) Forward stepwise selection, best subset selection, and forward stagewise regression.
(b) forward stepwise selection, forward stagewise regression and best subset selection.
(c) Best subset selection, forward stagewise regression and forward stepwise selection.
(d) Best subset selection, forward stepwise selection and forward stagewise regression.
Sol. b
7. (1 Mark) Choose the TRUE statements from the following: (Multiple correct choice)
(a) Ridge regression since it reduces the coefficients of all variables, makes the final fit a lot
more interpretable.
(b) Lasso regression since it doesn’t deal with a squared power is easier to optimize than
ridge regression.
(c) Ridge regression has a more stable optimization than lasso regression.
(d) Lasso regression is better suited for interpretability than ridge regression.
Sol. c, d
8. (2 Marks) Which of the following statements are TRUE? Let xi be the i-th datapoint in a
dataset of N points. Let v represent the first principal component of the dataset. (Multiple
answer questions)
PN
(a) v = arg max i=1 (v T xi )2 s.t. |v| = 1
PN
(b) v = arg min i=1 (v T xi )2 s.t. |v| = 1
(c) Scaling at the start of performing PCA is done just for better numerical stability and
computational benefits but plays no role in determining the final principal components
of a dataset.
(d) The resultant vectors obtained when performing PCA on a dataset can vary based on the
scale of the dataset.
Soln. A and D

You might also like