Learning the meso-scale organization of complex networks. This project aims to model and learn the organization of online social networks. We will combine mathematical models, inference, and domain knowledge from computational social sciences to obtain interpretable descriptions of the role groups of users play in the network. The expected outcomes are new mathematical models and computational methods that learn from data how to best decompose a complex network into building blocks and their int ....Learning the meso-scale organization of complex networks. This project aims to model and learn the organization of online social networks. We will combine mathematical models, inference, and domain knowledge from computational social sciences to obtain interpretable descriptions of the role groups of users play in the network. The expected outcomes are new mathematical models and computational methods that learn from data how to best decompose a complex network into building blocks and their interactions, linking connectivity to function. This should provide benefits to industries and policy makers interested in how information spreads in social media, including the critical questions of understanding the mechanisms contributing to political polarization and fragmentation.Read moreRead less
Ultra-sensitive 3D molecular assays using total body PET and deep learning. Recent advances in biomedical engineering have led to the development of Total Body Positron Emission Tomography (TB-PET), the most sensitive imaging device to date. Despite these impressive engineering advances, computational methods lag far behind and model-based approaches cannot deal with the complexity or volume of data these systems produce. We will develop new computational methods based on deep learning and stati ....Ultra-sensitive 3D molecular assays using total body PET and deep learning. Recent advances in biomedical engineering have led to the development of Total Body Positron Emission Tomography (TB-PET), the most sensitive imaging device to date. Despite these impressive engineering advances, computational methods lag far behind and model-based approaches cannot deal with the complexity or volume of data these systems produce. We will develop new computational methods based on deep learning and statistical methods that fully exploit the richness and complexity of the data generated by TB-PET, enabling 3D quantitative assays of molecular processes throughout the entire human body with unparalleled sensitivity. The technology we create will open up new capability for the study of complex physiological systems.Read moreRead less
Towards Generalisable and Unbiased Dynamic Recommender Systems. This project aims to develop the foundations, including models, methodology, and algorithms for building generalisable and unbiased dynamic recommender systems to facilitate intelligent decision-making, prompt contextualised and personalised strategic plans, and support context-aware action recourse. To ensure that fundamental principles, such as fairness and transparency, are respected, a set of algorithms and techniques are propos ....Towards Generalisable and Unbiased Dynamic Recommender Systems. This project aims to develop the foundations, including models, methodology, and algorithms for building generalisable and unbiased dynamic recommender systems to facilitate intelligent decision-making, prompt contextualised and personalised strategic plans, and support context-aware action recourse. To ensure that fundamental principles, such as fairness and transparency, are respected, a set of algorithms and techniques are proposed to develop recommender systems in a more responsible manner. The result of this project will not only maintain Australia's leadership in this frontier research area, but also serve as an excellent vehicle for the education and training of Australia's next generation of scholars and engineers.Read moreRead less
Uncertainty, Risk and Related Concepts in Machine Learning. Machine learning is the science of making sense of data. It does not and cannot remove all risk and uncertainty. This project proposes to study the foundations of how machine learning uses, represents and communicates risk and uncertainty. It aims to do so by finding new theoretical connections between diverse notions that have arisen in allied disciplines. These include risk, uncertainty, scoring rules and loss functions, divergences, ....Uncertainty, Risk and Related Concepts in Machine Learning. Machine learning is the science of making sense of data. It does not and cannot remove all risk and uncertainty. This project proposes to study the foundations of how machine learning uses, represents and communicates risk and uncertainty. It aims to do so by finding new theoretical connections between diverse notions that have arisen in allied disciplines. These include risk, uncertainty, scoring rules and loss functions, divergences, statistics and different ways of aggregating information. By building a more complete theoretical map it is expected that new machine learning methods will be developed, but more importantly that machine learning will be able to be better integrated into larger socio-technical systems.Read moreRead less
Trust-Oriented Data Analytics in Online Social Networks. Trust-oriented data analytics is essential in online social networks for reducing deceitful interactions and enhancing trust between users. This project aims to systematically devise innovative solutions by considering rich social contextual information as an important source of trust. The expected outcomes of this project include innovative solutions from a fundamental perspective to the challenges of context-aware trust propagation, trus ....Trust-Oriented Data Analytics in Online Social Networks. Trust-oriented data analytics is essential in online social networks for reducing deceitful interactions and enhancing trust between users. This project aims to systematically devise innovative solutions by considering rich social contextual information as an important source of trust. The expected outcomes of this project include innovative solutions from a fundamental perspective to the challenges of context-aware trust propagation, trust network searching/matching, and trustworthy/malicious user prediction in online social networks. This project is significant as it will advance the knowledge base for enabling a trustworthy social networking environment, benefiting billions of Australian and worldwide online social network users.Read moreRead less
Mitigating the Influence of Social Bots in Heterogeneous Social Networks. This project aims to mitigate the influence of social bots in dynamic and constantly changing social networks. Social bots can spread misinformation, manipulate public opinion, and compromise privacy and security. This project will use advanced algorithms to detect and neutralize the impact of social bots, improving the integrity and accuracy of information on social media. The expected outcomes include the development of ....Mitigating the Influence of Social Bots in Heterogeneous Social Networks. This project aims to mitigate the influence of social bots in dynamic and constantly changing social networks. Social bots can spread misinformation, manipulate public opinion, and compromise privacy and security. This project will use advanced algorithms to detect and neutralize the impact of social bots, improving the integrity and accuracy of information on social media. The expected outcomes include the development of a robust system for identifying and mitigating social bot influence, and the reduction of harmful content and misinformation on social media. The benefits of this project include a more trustworthy and secure social media environment, protection of individuals and organizations from malicious activities.Read moreRead less
Distributional Consequences of Mass-Market Higher Education in Business. Increased access to tertiary education has not been evaluated for its effects on the full spectrum of individuals served by the tertiary sector. Using longitudinal data on entire student populations at university business faculties, this project will provide the first Australian evidence on the trade-offs amongst the educational success of students with different levels of preparation that occur when those with poorer prep ....Distributional Consequences of Mass-Market Higher Education in Business. Increased access to tertiary education has not been evaluated for its effects on the full spectrum of individuals served by the tertiary sector. Using longitudinal data on entire student populations at university business faculties, this project will provide the first Australian evidence on the trade-offs amongst the educational success of students with different levels of preparation that occur when those with poorer preparation are added to classrooms. Short-term performance and medium-term attrition, a recent educational policy focus, will be evaluated. Theoretically grounded recommendations will result for undergraduate program design to suit a student population with varying levels of university preparation.Read moreRead less
Bayesian choice modelling. Discrete choice models are important as they provide tools to help understand choice processes of decision makers. It remains a challenge to specify models with covariance structures flexible enough to capture complex patterns of cross-substitution between choices while being able to capture heterogeneity present in individual behaviour. We will develop a Bayesian approach to choice modelling that uses covariance selection to overcome these problems. This will train re ....Bayesian choice modelling. Discrete choice models are important as they provide tools to help understand choice processes of decision makers. It remains a challenge to specify models with covariance structures flexible enough to capture complex patterns of cross-substitution between choices while being able to capture heterogeneity present in individual behaviour. We will develop a Bayesian approach to choice modelling that uses covariance selection to overcome these problems. This will train researchers and raise the profile of Australia in an active research area that is important in the social sciences; substantive applications will be in health economics, but developments will also be relevant to cognate areas of biostatistics, epidemiology, and ecology.Read moreRead less
Knowledge Graph-driven Software Vulnerability Risk Discovery and Assessment. This project aims to alleviate cyberattacks which are increasingly being crafted to attack software vulnerabilities and weaknesses by utilising advanced knowledge graphs and deep learning techniques. This project expects to construct an innovative software vulnerability knowledge graph and develop advanced graph-based algorithms and models. Expected outcomes of this project include the enhanced capacity to defend agains ....Knowledge Graph-driven Software Vulnerability Risk Discovery and Assessment. This project aims to alleviate cyberattacks which are increasingly being crafted to attack software vulnerabilities and weaknesses by utilising advanced knowledge graphs and deep learning techniques. This project expects to construct an innovative software vulnerability knowledge graph and develop advanced graph-based algorithms and models. Expected outcomes of this project include the enhanced capacity to defend against cyberattacks for both organisations and individuals in Australia and beyond, theory development in graph theory, refined graph neural network models and improved graph transfer learning algorithms.Read moreRead less
Feature Learning for High-dimensional Functional Time Series. This project aims to develop new methods and theories for common features on high-dimensional functional time series observed in empirical applications. The significance includes addressing a key gap in adaptive and efficient feature learning, improving forecasting accuracy and understanding forecasting-driven factors comprehensively for empirical data. Expected outcomes involve advances in big data theory and easy-to-implement algori ....Feature Learning for High-dimensional Functional Time Series. This project aims to develop new methods and theories for common features on high-dimensional functional time series observed in empirical applications. The significance includes addressing a key gap in adaptive and efficient feature learning, improving forecasting accuracy and understanding forecasting-driven factors comprehensively for empirical data. Expected outcomes involve advances in big data theory and easy-to-implement algorithms for applied researchers. This project benefits not only advanced manufacturing by finding optimal stopping time for wood panel compression, but also superior forecasting for mortality in demography, climate data in environmental science, asset returns in finance, and electricity consumption in economics. Read moreRead less