CDS503: Machine Learning ASSIGNMENT 1
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Academic Session: Semester 1, 2021-2020
ASSIGNMENT 1
Task Type
Assignment 1 is an individual assignment.
The student is required to choose one of the listed problems/data sets in Appendix A and propose the solution to the problem.
You are NOT required to perform an experiment on the chosen algorithms.
Assignment Description
Topic Selection: See Appendix A.
Based on the title chosen:
1. Data set background and characteristics. Study the data set chosen carefully. Based on the data set:
a. The background study of the data set. Perform the literature that covers (not limited to): (i) The usage of the dataset and the trend, including the proposed approaches (ii) Search for the literature for the past 5 years (2016 and above) that has used the
chosen dataset or similar ones. The required information includes the title, the authors, year, journal/conferences, and age number.
b. Report the class distribution of the given dataset (dataset in Appendix A).
c. Determine whether the data set is balanced or unbalanced. Give your justification and explain how this condition will affect the performance.
d. Feature correlations. Discuss and do a research on how the features correlations affect the performance of your classifier.
2. Pre-processing options. Discuss and select the suitable pre-processing options.
3. Model Evaluation Technique. Based on the size and characteristics of the data set, choose the suitable Model Evaluation techniques to be used in your machine learning evaluation. Explain the reason for using that option (hold-out, cross validation etc.).
4. Choice of the classifier. Choose one classifier that you have learned in the first half of the semester.
a. State the reasons of your choice based on the data set characteristics (and any others) and how these choices will affect the performance. You may justify your choice based on the literature done in 1(a).
b. Based on data set, select, and justify a suitable metric to evaluate the performance of your classification model (Confusion matrix, F1 Score etc.).
Report Requirement and Format
• A report must be prepared using Microsoft Word, font type Arial, size 12 in single line. Every chapter should start with a new page (Chapter 1.0 to 5.0, and references).
• A cover page should contains course name (including semester and year), assignment title, name, matrix no and dataset title.
• Table of Contents
1.0 Dataset Background
2.0 Pre-processing options
3.0 Model Evaluation Technique
4.0 Choice of the classifier
5.0 Conclusion
References
Note: You may create additional subsection as deemed necessary.
Report Submission Instruction
• Submit soft copy (zip/rar to eLearn@USM).
• The zip/rar package must be named according to the following notation:
CDS503_Assignment 1_Name_MatrixNo_TitleNo.
Assignment Evaluation
This assignment will be graded (A to F scale).
IMPORTANT: Students who copied or plagiarized other’s work or let their work be copied or plagiarized will be given an F grade. The student may be barred from sitting for final exam and
reported to the university’s disciplinary board.
2021-12-13