Scribd
Upload
42 views

Assignment 2

This document contains 6 multiple choice and short answer questions about machine learning concepts like linear regression, normalization, k-nearest neighbors algorithm, k-means clustering algorithm, data smoothing, perceptrons, feedforward neural networks, and data visualization best practices. It asks the reader to perform tasks like creating models, explaining choices of algorithms, showing step-by-step working, and identifying issues with a sample pie chart visualization.

Uploaded by

Fanny Cheung
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
42 views

Assignment 2

This document contains 6 multiple choice and short answer questions about machine learning concepts like linear regression, normalization, k-nearest neighbors algorithm, k-means clustering algorithm, data smoothing, perceptrons, feedforward neural networks, and data visualization best practices. It asks the reader to perform tasks like creating models, explaining choices of algorithms, showing step-by-step working, and identifying issues with a sample pie chart visualization.

Uploaded by

Fanny Cheung
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Question 1. You are given the following data about the electricity bills of an apartment.

(20
marks)

1.1 Please create a linear regression using the average monthly temperature to predict the
electricity bill without normalizing the data. (10 marks)

1.2 If the above is normalized by decimal scaling, do you think the answer obtained will be
different? Please give reasons to support your argument. No calculation is required. (10 marks)

Question 2. You are given the following data about the interview score sheet of some job
applicants. (15 marks)

2.1 To predict the interview result for Eva, is it better to apply K-NN algorithm or K-means
algorithm? Explain your reason. (5 marks)

2.2 With the answer you have provided in 2.1, run that algorithm with K = 3. Show your steps.
(10 marks)
Question 3. Answer the following questions. (25 marks)

3.1 Describe the motivation of data normalization. (5 marks)

3.2 Suppose that you have the following data: [100, 200, 300, 400], normalize the data by min-
max normalization by setting new min to 5 and new max to 8. (10 marks)

3.3 Suppose we have the following values: [4, 34, 31, 30, 21, 31, 21, 25, 22, 28, 10, 15]. Use
smoothing by bin means to smooth the above data using equal-depth binning with 3 bins. Show
your steps. (10 marks)

Question 4. Consider the following dataset, can perceptron separate the points of y=1 and y=-1?
Justify your answer. (5 marks)
Question 5. Consider the following feed-forward neural network. The model weights are shown
2
on the edges. Suppose we use the sigmoid function (i.e., f ( x )= ) as the activation
1+exp ⁡(−2 x )
function in this network. (15 marks)

Given a sample with x1 = 0 and x2 =0, what are the output values of each node.
Node Output Value
h1(1)
h2(1)
h1(2)
h2(2)
o

Question 6. Although the Pie chart is a common visualization, it is often misused. Given the pie
chart below, list at least three design issues of this pie chart. (10 marks)

In addition, name another visualization you would like to choose to visualize the application
category percentage (10 marks)

You might also like