There is a more recent version of this academic item available.

Overview

Deep learning is currently the central machine learning method fuelling the artificial intelligence revolution. In this subject you learn how to apply deep learning algorithms to solve real-world problems. This subject does not assume you have previous machine learning experience, therefore it starts teaching deep learning at a very introductory … For more content click the Read more button below.

Portfolio

Science, Health & Engineering (Pre 2022)

Subject coordinator

Zhen He

Subject type

Postgraduate

Year level

Year Level 5 - Masters

AQF level

Level 9 - Masters Degree

Available as elective

Yes

Available to study abroad / exchange students

Yes

Capstone subject

No

Academic progress review - Schedule A subject

No

Subject instances

To view instance specific details which include - Assessments, Class requirements and Subject instance coordinators - please select your preferred instance via the drop-down menu at the top right-hand side of this page.

Learning resources

Career ready

Work based learning (placement):No

Subject intended learning outcomes

On successful completion you will be able to:
1.
Analyse the advantages and disadvantages of using traditional machine learning algorithms versus deep learning algorithms for solving industry problems.
2.
Analyse a given industry problem and then recommend which deep learning algorithm best solves the problem. This requires the student to know which best learning algorithm is most suitable to solve each type of problem.
3.
Write/produce Pytorch code to implement deep learning algorithms to solve computer vision and natural language processing problems.
4.
Propose deployment and maintenance strategies for deep learning production systems in the cloud based on the analysis from the machine learning requirements of a business.
5.
Given a real-world problem where labels are scarce, investigate different deep learning approaches to tackle this problem.
6.
Design the architecture of a generative adversarial network that can generate realistic data.