ECE5179 - Neural networks and deep learning

Overview

This unit introduces fundamentals of deep learning and how it can solve problems in many areas such as image classification, filter design and natural language processing. Neural networks are first described and how training can be achieved with backpropagation. Various forms of deep neural networks are developed such as multilayer perceptrons, convolution neural networks and recurrent neural networks. Deep reinforcement learning is discussed. The mathematics of stochastic optimisation is used to interpret and understand the behaviour and training of these networks. Programming approaches are discussed for training and deploying neural networks. Deep learning technologies and design examples are discussed in areas such as robotics, driverless cars, personal cognitive assistants and mastering of games such as GO.

Offerings

S2-01-CLAYTON-FLEXIBLE

Requisites

Prohibition

Rules

Enrolment Rule

Contacts

Chief Examiner(s)

Associate Professor Mehrtash Tafazzoli Harandi

Unit Coordinator(s)

Associate Professor Mehrtash Tafazzoli Harandi

Learning outcomes

On successful completion of this unit, you should be able to:

Describe concepts and fundamentals of deep learning such as the backpropagation algorithm and adversarial learning.

Appraise various forms of deep neural networks such as multilayer perceptrons, convolution neural networks and recurrent neural networks.

Interpret and apply the mathematics of deep learning such as stochastic optimisation.

Design deep learning solutions to problems in computer vision, natural language processing and signal processing. Examples are image classification, object detection, sequence modelling and filter design.

Design and synthesise the training and deployment of neural networks using a high-level programming language.

Critically assess sources of information and contents of scientific publications and choose relevant information

Teaching approach

Problem-based learning

Active learning

Online learning

Assessment summary

Continuous assessment: 50%

Final assessment: 50%

This unit contains hurdle requirements that you must achieve to be able to pass the unit. You are required to achieve at least 45% in the total continuous assessment component and at least 45% in the final assessment component. The consequence of not achieving a hurdle requirement is a fail grade (NH) and a maximum mark of 45 for the unit.

The assessments of this unit are designed to demonstrate the achievement of the advanced learning outcomes and standards expected of Master’s level coursework.

Assessment

1 - Assignments

2 - Quizzes

3 - Project

4 - Lab assessments

5 - Final assessment

Scheduled and non-scheduled teaching activities

Studio activities

Workshops

Workload requirements

Workload

The minimum total expected workload to achieve the learning outcomes for this unit is 144 hours per semester typically comprising a mixture of 3-6 hours of scheduled learning activities and 6-9 hours of independent study per week. Scheduled activities may include a combination of teacher-directed learning, peer-directed learning and online engagement. Independent study may include associated readings, assessment and preparation for scheduled activities.

Learning resources

Required resources

Recommended resources

Technology resources

Availability in areas of study

E6014 Master of Engineering - Specialisation: Power systems engineering
E6017 Master of Advanced Engineering - Specialisation: Robotic construction engineering
E6017 Master of Advanced Engineering - Specialisation: Smart manufacturing engineering