CSIT970 Security Essentials AUT Assignment Question 2025 | UOW

CSIT970 Assignment Question:

Q1: This question is about Privacy Preserving Techniques for Machine Learning. (10 marks) 

A.Assume that Alice's salary is $70,000, Bob's salary is $50,000, and Eve's salary is $95,000. They want to calculate the average salary but don't want to reveal their salaries to others. Explain how they can achieve this secure three-party computation using (3,3) secret sharing. Show your work to get a full mark. (3 marks) 

B.In this task, you compare the privacy preserving techniques for machine learning mentioned in Week 8's lecture, namely Multi-Party Computation (MPC), Homomorphic Encryption (HE), and Differential Privacy (DP). Gather relevant information about MPC, HE and DP as much as you can and write a report about them including the following contents: 

• Summary 
• Overview of MPC, HE, and DP - how they work and examples 
• What are their strengths and limitations 
• Provide a comparative summary table 
• Your thoughts on their techniques in ML/AI 
• References 

*There are many websites/articles comparing the those techniques, available from the Internet. You can refer to them, but proper citations must be made. (Failure to do so may be regarded as plagiarism.)

To get full 7 marks for this question, you need to satisfy the following requirements. 

Requirements: 

• The length of your report should be 1500 words excluding references.
• Your report should be single-spaced with font size 12 and be typed using MS Word. 
• Name your report as Q1.docx (or Q1.doc). 

Your report will be marked based on presentation, clarity, structure and succinctness of the report contents. 

Q2: (10 marks)Assume that the hacking pattern appears in the log file as client's request that consists of URI and protocol.