Developing causality-based techniques towards responsible AI

About this project

Artificial Intelligence (AI) is becoming increasingly prevalent and affects people’s everyday lives. This established fact highlights the urgent need for new techniques that push AI models beyond mere prediction accuracy, with responsible AI being key to the current deployment of AI systems. Responsible AI models must be free of bias, performing consistently across all population subgroups, and must not unfairly disadvantage any group. They must also provide reliable, evidence-based explanations to assist users in making informed decisions, particularly in high-stakes scenarios. In contrast, irresponsible AI models that exhibit bias and fail to explain their predictions often lead to undesirable outcomes, such as failed COVID-19 forecasting that results in misinformed policy-making, discriminatory recruitment practices, and incorrect cancer treatment recommendations.

Causality is essential for building responsible AI models. Much like human beings, being responsible requires an understanding of the underlying causal mechanisms that govern how and why things happen. The challenge, however, is enabling computers to learn and apply the correct causal relationships in real-world scenarios and develop causal reasoning capabilities.

This project aims to achieve the following objectives:

1.     Develop novel methods to discover causal relationships within models, aided by data and domain knowledge, and use these relationships to prevent and mitigate biases in the model's predictions.

2.     Design causal explanation methods that provide logical, evidence-based justifications for a model’s predictions, helping users make fully informed decisions when using AI systems.

Applicants are required to demonstrate excellent programming skills. Scientific publications are desirable.


What you’ll do 

You will study in a supportive environment within the Data Analytics Group of the Industrial AI Research Centre. The group's research has been supported by numerous Australian Research Council Discovery Projects and industry funding (http://4llab.net/projects.html). Over the past five years, the group has successfully supervised 15 PhD students to completion and has more than a decade of research experience in data mining and machine learning. The group includes research fellows and PhD students working on related topics, and weekly seminars showcase current progress in various areas of data mining and machine learning. This supportive environment fosters success for new research fellows and PhD students.

You will be expected to conduct the research outlined in the project overview and engage in implementation.

You will also be expected to attend at least one international conference during their PhD studies.


Where you’ll be based 

The student will study in a supportive environment within the Data Analytics Group of the Industrial AI Research Centre. The group's research has been supported by numerous Australian Research Council Discovery Projects and industry funding (http://4llab.net/projects.html). Over the past five years, the group has successfully supervised 15 PhD students to completion and has more than a decade of research experience in data mining and machine learning. The group includes research fellows and PhD students working on related topics, and weekly seminars showcase current progress in various areas of data mining and machine learning. This supportive environment fosters success for new research fellows and PhD students. Financial Support

This project is funded for reasonable research expenses. A fee offset for the standard term of the program is available to Australian and New Zealand citizens, and permanent residents of Australia, including permanent humanitarian visa holders.  Additionally, any Australian Aboriginal and/or Torres Strait Islander applicant who holds an offer of admission without a living allowance will be eligible for the Aboriginal Enterprise Research Scholarship. This scholarship is to the value of $52,352 per annum (2025 rate). Any Aboriginal Enterprise Research Scholarship recipient will also receive a fee waiver. Where an international applicant holds an external scholarship or sponsorship a full or partial fee waiver may apply in some circumstances for exceptional applicants.  Other international applicants will be required to pay full tuition fees of approximately $44,300 per annum (2025 rates).  


Eligibility and Selection 

This project is open to applications from both Domestic and International applicants.

Applicants must meet the eligibility criteria for entrance into a PhD. Additionally, applicants must meet the project selection criteria: 

• Demonstrated excellent programming skills.
• Scientific publications are desirable.

All applications that meet the eligibility and selection criteria will be considered for this project. A merit selection process will be used to determine the successful candidate.

The successful applicant is expected to study full-time, and to be based at our Mawson Lakes campus in the north of Adelaide. Note that international students on a student visa will need to study full-time.


Essential Dates

Applicants are expected to start in a timely fashion upon receipt of an offer. Extended deferral periods are not available. Applications close on 31 January 2025.