Advanced Machine Learning with Bilevel Optimization. There is an urgent need to develop a new machine learning (ML) paradigm that can overcome data-privacy and model-size constraints in real-world applications. This project aims to develop an advanced paradigm of ML with bilevel optimisation, called bilevel ML. A theoretically-guaranteed fast approximate solver and a new fuzzy bilevel learning framework will be developed to achieve the aim in complex situations; a methodology to transfer knowled ....Advanced Machine Learning with Bilevel Optimization. There is an urgent need to develop a new machine learning (ML) paradigm that can overcome data-privacy and model-size constraints in real-world applications. This project aims to develop an advanced paradigm of ML with bilevel optimisation, called bilevel ML. A theoretically-guaranteed fast approximate solver and a new fuzzy bilevel learning framework will be developed to achieve the aim in complex situations; a methodology to transfer knowledge and an approach to fast-adapt bilevel optimization solutions when required computing resources change. The anticipated outcomes should significantly improve the reliability of ML with benefits for safety learning and computing resource optimisation in ML-based data analytics.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
A Novel Framework for Optimised Ensemble Classifier. The project aims to develop a novel framework for creating an optimised ensemble classifier that will improve data analysis and the accuracy of many real-world applications such as document analysis, robotics and medical diagnosis. The project plans to develop and investigate novel methods for generating diverse training environment layers, base classifiers and fusion of classifiers. It also plans to design a multi-objective evolutionary algor ....A Novel Framework for Optimised Ensemble Classifier. The project aims to develop a novel framework for creating an optimised ensemble classifier that will improve data analysis and the accuracy of many real-world applications such as document analysis, robotics and medical diagnosis. The project plans to develop and investigate novel methods for generating diverse training environment layers, base classifiers and fusion of classifiers. It also plans to design a multi-objective evolutionary algorithm-based search obtain the optimal number of layers, clusters and base classifiers. The expected outcomes of the proposed framework are advances in classifier learning. The final outcome may be novel methods which will bring in diversity during the learning of the base classifiers and provide an optimal ensemble classifier for real-world applications.Read moreRead less
Mechanisms of nerve fibre guidance by molecular gradients. Brain wiring is crucial for brain function. The project will investigate the basic principles underlying the development of brain wiring, using both experiments and mathematical models. This will lead a predictive model of how wiring develops, both in normal and abnormal situations.
An automated system for the analysis of road safety and conditions. This project aims to develop an automated system for the analysis of road safety and conditions. Digital video road data is collected over every state road in Queensland annually, and has the potential to provide a range of value-added products for safer roads. This project will develop deep learning based neural network techniques which can learn and classify roadside objects so that video data can be automatically analysed all ....An automated system for the analysis of road safety and conditions. This project aims to develop an automated system for the analysis of road safety and conditions. Digital video road data is collected over every state road in Queensland annually, and has the potential to provide a range of value-added products for safer roads. This project will develop deep learning based neural network techniques which can learn and classify roadside objects so that video data can be automatically analysed allowing the estimation of the proximity of objects for road safety and rating. The expected outcome will be new identification techniques and software which can be incorporated with road data collection systems.Read moreRead less
Deep Learning Architecture with Context Adaptive Features for Image Parsing. This project aims to develop a novel deep learning network architecture with contextual adaptive features for image parsing that can improve the object detection accuracy in real-world applications. A number of innovative methods for deep learning, contextual features and network parameter selection will be developed and investigated. The impact of the proposed architecture and features will be improved object-detection ....Deep Learning Architecture with Context Adaptive Features for Image Parsing. This project aims to develop a novel deep learning network architecture with contextual adaptive features for image parsing that can improve the object detection accuracy in real-world applications. A number of innovative methods for deep learning, contextual features and network parameter selection will be developed and investigated. The impact of the proposed architecture and features will be improved object-detection accuracy and advances in deep learning network architecture for image parsing. The intended outcomes are deep learning network architecture, contextual feature extraction techniques and network parameter optimisation techniques for image parsing.Read moreRead less
Bio-inspired Computing for Problems with Dynamically Changing Constraints. The aim of this project is to design bio-inspired computing methods for dynamically changing environments. Dynamic problems arise frequently in the areas of engineering, logistics, and manufacturing. Such problems are usually subject to a large set of constraints that change over time due to changes in resources. Algorithms that can deal with such dynamic changes would benefit decision-makers. The project aims to provide ....Bio-inspired Computing for Problems with Dynamically Changing Constraints. The aim of this project is to design bio-inspired computing methods for dynamically changing environments. Dynamic problems arise frequently in the areas of engineering, logistics, and manufacturing. Such problems are usually subject to a large set of constraints that change over time due to changes in resources. Algorithms that can deal with such dynamic changes would benefit decision-makers. The project aims to provide a foundational theory as the basis for the design of bio-inspired algorithms dealing with dynamically changing constraints and provide approaches for dealing with important industrial problems.Read moreRead less
Bio-inspired Computing for Problems with Chance Constraints. Bio-inspired algorithms have successfully been applied to a wide range of optimisation problems. Uncertainties in real-world applications can lead to critical failures of production schedules or safe critical systems. Chance constraints model such uncertainties and allow to limit the possibility of such failures. This future fellowship builds up the area of bio-inspired computing for problems with chance constraints. It develops high ....Bio-inspired Computing for Problems with Chance Constraints. Bio-inspired algorithms have successfully been applied to a wide range of optimisation problems. Uncertainties in real-world applications can lead to critical failures of production schedules or safe critical systems. Chance constraints model such uncertainties and allow to limit the possibility of such failures. This future fellowship builds up the area of bio-inspired computing for problems with chance constraints. It develops high performing bio-inspired algorithms for stochastic problems where the constraints can only be violated with a small probability. The outcomes will lead to more effective and reliable optimisation methods for complex planning processes in areas of national priority such as mining and manufacturing.Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE140100017
Funder
Australian Research Council
Funding Amount
$394,800.00
Summary
Adaptive Optimisation of Complex Combinatorial Problems. One of the most common problems faced by planners, whether in industry or government, is optimisation, finding the optimal solution to a problem. Even a one per cent improvement in a solution can make a difference of millions of dollars in some cases. Traditionally optimisation problems are solved by analytic means or exact optimisation methods. Today, however, many optimisation problems involve complex combinatorial systems that make such ....Adaptive Optimisation of Complex Combinatorial Problems. One of the most common problems faced by planners, whether in industry or government, is optimisation, finding the optimal solution to a problem. Even a one per cent improvement in a solution can make a difference of millions of dollars in some cases. Traditionally optimisation problems are solved by analytic means or exact optimisation methods. Today, however, many optimisation problems involve complex combinatorial systems that make such traditional approaches unsuitable or intractable. This project aims to assist researchers and practitioners in solving complex combinatorial optimisation problems by adapting the optimisation strategy to the problem being solved, based on problem features such as search space difficulty. Read moreRead less
A Novel Automatic Neural Network Feature Extractor. This project aims to study feature extraction abilities of convolutional as well as traditional neural networks and develop a generic feature extractor which can be applied to wide variety of real-world image and non-image data. New concepts for automatic feature extraction, feature explanation, hybrid evolutionary algorithms and non-iterative ensemble learning will be introduced and evaluated. The expected outcomes are a generic feature extrac ....A Novel Automatic Neural Network Feature Extractor. This project aims to study feature extraction abilities of convolutional as well as traditional neural networks and develop a generic feature extractor which can be applied to wide variety of real-world image and non-image data. New concepts for automatic feature extraction, feature explanation, hybrid evolutionary algorithms and non-iterative ensemble learning will be introduced and evaluated. The expected outcomes are a generic feature extractor for automatically extracting features, an optimiser for finding optimal parameters and non-iterative ensemble learning technique for classification of features into classes. The impact of this project will be automatic feature extractors and classifiers for real-world applications.Read moreRead less