COMP90049 Introduction to Machine Learning
University of Melbourne · Level 9 · 12.5 points · Runs in 2026: Semester 1, Semester 2.
- Final exam 50%
- 30% group work
Is COMP90049 hard?
COMP90049: Harder. A graduate subject, so it isn't on the easy and hard lists; this is how it compares with Bachelor of Science's subjects. Makes it harder: Reviewers say it’s hard (4 reviews), Reviewers say it takes a lot of time, You must pass the exam on its own.
What you learn
Machine learning basics: cleaning data, classification, regression, clustering and the maths behind them.
After it you can
- Derive ML models from first principles
- Build and evaluate ML systems for real problems
- Choose the right model for a problem
What you need first
COMP20003 or COMP20007 or ENGR30004 or ENGR30003
- COMP20003 Algorithms and Data Structures
- COMP20007 Design of Algorithms
- COMP90038 Algorithms and Complexity
- ENGR30004 Numerical Algorithms in Engineering
Can’t be taken together with
- COMP30027 Machine Learning
- COMP30018
How it’s assessed
Projects 50%, Final exam 50%
Student reviews and ratings
Summarised from 4 students on StudentVIP (2017–2019)
- An introduction to machine learning and information retrieval. Reviewers found it interesting and useful for data science.
- Lectures are compact and dense, so expect to invest a lot of your own time. Tutorials were helpful.
- Projects are self-directed with loose instructions, and the final exam is tough to finish in time.
A summary of public reviews, not scored here.
Taken this subject? Rate it anonymously (difficulty, workload, marking, would you recommend it) and leave a tip for next year's students. Rate COMP90049
Quick answers
- When is COMP90049 offered?
- Runs in 2026: Semester 1, Semester 2.
- What do I need before taking COMP90049?
- Needs COMP20003 or COMP20007 or ENGR30004 or ENGR30003.
- Does COMP90049 have a final exam?
- Yes, the final exam is 50% of the mark. You must pass the exam on its own. Assessment: Projects 50%, Final exam 50%.
- Is COMP90049 hard?
- From 4 student reviews on StudentVIP: An introduction to machine learning and information retrieval. Reviewers found it interesting and useful for data science.
- What do students say about COMP90049?
- An introduction to machine learning and information retrieval. Reviewers found it interesting and useful for data science. Lectures are compact and dense, so expect to invest a lot of your own time. Tutorials were helpful. Projects are self-directed with loose instructions, and the final exam is tough to finish in time.
- Can I take COMP90049 as breadth?
- No, it isn’t listed as breadth for any course.
- Can I rate COMP90049?
- Yes. If you've taken it, rate it anonymously on this page (difficulty, workload, marking, would you recommend it) and leave a tip for next year's students. Averages show once 3 students have rated it; tips are checked before they appear.
Source: University of Melbourne Handbook 2026
Always confirm in the official Handbook.
PickMySubjects is a free, unofficial planner made by a UniMelb graduate.