ggroundIt

Past year questions

Machine Learning Techniques (BSCS2007) PYQs

57 verified past-year questions tagged to real concepts, week by week. A few samples below — the full bank, filterable by type and difficulty, is in the app.

Distribution by week

Week 1
22
Week 2
9
Week 3
12
Week 4
14

Sample questions

mcq·hard·pyq4

Let x = [x₁, x₂]ᵀ and y = [y₁, y₂]ᵀ be two vectors in ℝ². Let k₁ and k₂ be two functions defined from ℝ² × ℝ² → ℝ:

k₁(x, y) = x₁y₁ + x₂y₂ + x₁x₂y₁y₂ k₂(x, y) = x₁y₁ + x₂y₂ + x₁x₂y₁y₂ + 1

Which of the following statements is true?

A.

Both k₁ and k₂ are valid kernels.

B.

k₁ is a valid kernel, but k₂ is not a valid kernel.

C.

k₂ is a valid kernel, but k₁ is not a valid kernel.

D.

Neither k₁ nor k₂ is a valid kernel.

msq·medium·pyq4

Consider a covariance matrix C for a mean-centered dataset in ℝ³. After performing standard PCA, the three principal components turn out to be:

[1, 0, 0]ᵀ, [0, 1, 0]ᵀ, [0, 0, 1]ᵀ

Which of the following statements are true? You can assume that C is not the zero matrix.

A.

C is a diagonal matrix.

B.

The diagonal entries of C are non-negative.

C.

The diagonal entries of C are strictly greater than zero.

D.

C has to be the identity matrix.

E.

C is a matrix of the form kI, where k > 0 and I is the identity matrix.

nat·hard·pyq4

Kernel PCA with a polynomial kernel of degree 3 is run on a dataset of 100 data-points. Let φ be the transformation corresponding to this kernel. If the largest eigenvalue of the centered kernel matrix is 10, find the largest eigenvalue of the covariance matrix of the transformed dataset, where φ is used to effect the transformation.

Unofficial, student-built, not affiliated with IIT Madras.