[ TPR2251 ] Quiz 3 - Set 1

Which of the following techniques would perform better for reducing dimensions of a data set?

• Removing columns which have too many missing values

• Removing columns which have high variance in data

• Removing columns with dissimilar data trends

Which of the following can be the first 2 principal components after applying PCA?

I. ( 0.5, 0.5, 0.5, 0.5 ) and ( 0.71, 0.71, 0, 0 )

II. ( 0.5, 0.5, 0.5, 0.5 ) and ( 0, 0, -0.71, -0.71 )

III. ( 0.5, 0.5, 0.5, 0.5 ) and ( 0.5, 0.5, -0.5, -0.5 )

IV. ( 0.5, 0.5, 0.5, 0.5 ) and ( -0.5, -0.5, 0.5, 0.5 )

• I and II

• I and III

• II and IV

• III and IV

Imagine you are dealing with 10 class classification problem and you want to know that at most how many discriminant vectors can be produced by LDA. What is the correct answer?

• 10

• 9

• 8

• 7

The given dataset consists of images of “Hoover Tower” and some other towers. Now, you want to use PCA and the nearest neighbour method to build a classifier that predicts whether new image depicts “Hoover tower” or not.

In order to get reasonable performance, what pre-processing steps will be required on these images?

I. Align the towers in the same position in the image

II. Scale or crop all images to the same size

III. Apply image filtering

• I and II

• I and III

• II and III

• I, II, and III

Consider an application of the k-means clustering algorithm to the one dimensional data set D = { 0, 1, 5, 8, 14, 16 } for k = 3 clusters.

Start with the three clusters means: m1 = 2, m2 = 6 and m3 = 9. What are the values of the means at the next iteration?

• m1 = 0.5; m2 = 5; m3 = 12.67

• m1 = 2; m2 = 5; m3 = 9

• m1 = 0.5; m2 = 6; m3 = 8.53

• m1 = 0.5; m2 = 2.5; m3 = 5.64

Let the following be the one-dimensional data classified in three classes:

C1 = { 2, 6, 7, 8, 10, 13 }

C2 = { 0, 1, 5, 14 }

C3 = { 3, 4, 9, 11, 12 }

Classify a sample x = 5.9 using the k-nearest neighbourhood (k-NN) algorithm with k = 5.

• C1

• C2

• C3

• Can't say

Consider a neural network consisting of four inputs with the respective weights as shown in the figure.

What is the output of the neuron Y, if a symmetric hard limit transfer function with threshold = 0 is used?

• 15

• 14

• 1

• 0

What will be a classification diagram look like for a support vector machine that has overfit the data?

• There will be very few data points as support vectors

• Most of the data points would be support vectors

• The diagram will be very sparse

• There will be many decision boundaries

Which of the following is TRUE for the approach to combine a set of data points in agglomerative clustering?

• In single linkage, we define the distance between two clusters to be the minimum distance between any single data point in the first cluster and any single data point in the second cluster.

• In complete linkage, we define the distance between two clusters to be the mode distance between any single data point in the first cluster and any single data point in the second cluster.

• In average linkage, we define the distance between two clusters to be the median distance between data points in the first cluster and data points in the second cluster.

• The single linkage and complete linkage combines two data points based on Euclidean Distance.

Determine the point of learning an optimal machine learning model for pattern recognition task.

• Point A

• Point B

• Point C

• Point B or C