Causal Knowledge-Empowered Adaptive Federated Learning. Federated learning tools are a promising framework for collaborative machine learning (ML) that also maintain data privacy; however, their ability to model heterogeneous data remains a key challenge. This project aims to develop a new learning scheme for coordinated training of ML models that successfully bridges variable data distributions. The framework proposed will be the first globally that can use causal knowledge to 1) handle data he ....Causal Knowledge-Empowered Adaptive Federated Learning. Federated learning tools are a promising framework for collaborative machine learning (ML) that also maintain data privacy; however, their ability to model heterogeneous data remains a key challenge. This project aims to develop a new learning scheme for coordinated training of ML models that successfully bridges variable data distributions. The framework proposed will be the first globally that can use causal knowledge to 1) handle data heterogeneity across devices and 2) address the real-world challenges when only a subset of devices have labelled data. Expected outcomes and benefits include the theoretical underpinnings and algorithms of causality-based collaborative training of ML models while better preserving the users’ data privacy.Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE240101089
Funder
Australian Research Council
Funding Amount
$436,847.00
Summary
Trustworthy Hypothesis Transfer Learning. It is urgent to develop a new hypothesis transfer learning scheme that can overcome potential risks when finetuning unreliable large-scale pre-trained models. This project aims to develop an advanced and reliable scheme of hypothesis transfer learning, called Trustworthy Hypothesis Transfer Learning (TrustHTL). A new theoretically guaranteed heterogeneous hypothesis transfer learning framework will be developed to handle heterogeneous situations; a metho ....Trustworthy Hypothesis Transfer Learning. It is urgent to develop a new hypothesis transfer learning scheme that can overcome potential risks when finetuning unreliable large-scale pre-trained models. This project aims to develop an advanced and reliable scheme of hypothesis transfer learning, called Trustworthy Hypothesis Transfer Learning (TrustHTL). A new theoretically guaranteed heterogeneous hypothesis transfer learning framework will be developed to handle heterogeneous situations; a methodology to disinherit risks of pre-trained models and a new fuzzy relation based distributional discrepancy in heterogeneous transfer learning scenarios. The outcomes should significantly improve the reliability of machine learning with benefits for safety learning in data analytics.Read moreRead less
Exploiting Geometries of Learning for Fast, Adaptive and Robust AI. This project aims to uniquely exploit geometric manifolds in deep learning to advance the frontier of Artificial Intelligence (AI) research and applications in cybersecurity and general cognitive tasks. It expects to develop new theories, algorithms, tools, and technologies for machine learning systems that are fast, adaptive, lifelong and robust, even with limited supervision. Expected outcomes will enhance Australia's capabili ....Exploiting Geometries of Learning for Fast, Adaptive and Robust AI. This project aims to uniquely exploit geometric manifolds in deep learning to advance the frontier of Artificial Intelligence (AI) research and applications in cybersecurity and general cognitive tasks. It expects to develop new theories, algorithms, tools, and technologies for machine learning systems that are fast, adaptive, lifelong and robust, even with limited supervision. Expected outcomes will enhance Australia's capability and competitiveness in AI, and deliver robust and trustworthy learning technology. The project should provide significant benefits not only in advancing scientific and translational knowledge but also in accelerating AI innovations, safeguarding cyberspace, and reducing the burden on defence expenses in Australia.Read moreRead less
Self-supervised feature learning for rapid processing of marine imagery. Fast and reliable quantitative estimates of marine environmental health are needed for scientific studies, design and management of protected areas, and regulatory compliance of industrial activity in the ocean. Australia is collecting seafloor images at increasing rates but expert annotations are not keeping up, meaning that typical machine learning approaches struggle. This project will develop self-supervised techniques ....Self-supervised feature learning for rapid processing of marine imagery. Fast and reliable quantitative estimates of marine environmental health are needed for scientific studies, design and management of protected areas, and regulatory compliance of industrial activity in the ocean. Australia is collecting seafloor images at increasing rates but expert annotations are not keeping up, meaning that typical machine learning approaches struggle. This project will develop self-supervised techniques that use large amounts of unlabeled data to enhance performance. Our design takes advantage of additional information available for marine imagery such as geolocation and remote sensing context. We will explore how these representations can guide additional sampling and improve performance in classification tasks.Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE240100144
Funder
Australian Research Council
Funding Amount
$444,447.00
Summary
Universal Model Selection Criteria for Scientific Machine Learning. This project aims to develop provably reliable universal model selection criteria to facilitate trustworthy scientific machine learning. Combining stochastic methods with an innovative geometric approach to basic statistical principles, this project expects to characterise, combine, and refine the most successful heuristics for designing and training huge models, such as deep neural networks, into a cohesive theoretical framewor ....Universal Model Selection Criteria for Scientific Machine Learning. This project aims to develop provably reliable universal model selection criteria to facilitate trustworthy scientific machine learning. Combining stochastic methods with an innovative geometric approach to basic statistical principles, this project expects to characterise, combine, and refine the most successful heuristics for designing and training huge models, such as deep neural networks, into a cohesive theoretical framework. The expected outcomes include a general toolkit for assisting neural network design at the forefront of scientific applications. This should significantly improve the quality of scientific predictions by facilitating confident adoption of deep learning methods into the pantheon of trustworthy modeling techniques. Read moreRead less