ST3189 Machine learning
Tuesday 17 December 2023
The assessment will be a closed-book take-home online assessment within a 3-hour
window. The expected time/effort to answer all questions is 2 hours.
Candidates should answer all FOUR questions. All questions carry equal marks.
A table of common distributions is provided after the final question of this paper.
You should complete this paper using pen and paper. Please use BLACK INK only.
Please ensure that your candidate number is written clearly at the top of each page. Please
do not write your name anywhere on your submission.
Workings should be submitted for all questions requiring calculations(if any). Any
necessary assumptions introduced in answering a question are to be stated.
You may use any calculator for any appropriate calculations, but you may not use any
computer software to obtain solutions. Credit will only be given if all workings are shown.
If you think there is any information missing or any error in any question, then you should
indicate this but proceed to answer the question stating any assumptions you have made.
Question 1
a) Ensemble models are a machine learning approach to combine multiple other
models in the prediction process. These models are referred to as base estimators.
Explain how ensemble methods overcome most of the technical challenges in
single estimators.
(5 marks)
b) Explain what a bagging model is and a boosting model and compare these two
techniques with examples.
(10 marks)
c) Briefly explain each of the below performance matrices in supervised learning
I. Accuracy
II. Precision
III. Recall
IV. F1 score
(10 marks)
d) What is an imbalanced dataset and the issues of imbalanced data usage in model
training?
(5 marks)
e) Explain the practical issues associated with accuracy with examples and the other
performance measures we can use for model validations and how each measure
can overcome the issues you identified.
(10 marks)
Question 2
a) Briefly explain the difference between classification modeling and regression
modeling.
(5 marks)
b) Explain the reasons with examples why the performance measures used in
classification measures do not fit for regression model validation.
(10 marks)
c) Briefly explain each of the below performance measures used in regression analysis
I. SSE
II. MSE
III. RMSE
IV. R2
(10 marks)
d) Table 1 shows the results obtained by a linear regression model against the testing
dataset. Calculate SSE, MSE, and RMSE of the model.
(15 marks)
Table 1: Model validation results
Test value Predicted value
68 71
77 76
81 76
82 81
88 86
90 92
Question 3
You are working on a binary classification problem with classes: Positive (P) and Negative (N). The
dataset is split into three groups after applying a decision split.
Split Group Positive (P) Negative (N) Total Count
Group 1 40 10 50
Group 2 10 30 40
Group 3 10 0 10
(20 marks)
