Discovery Early Career Researcher Award - Grant ID: DE240100006
Funder
Australian Research Council
Funding Amount
$444,847.00
Summary
Robust Derivative-Free Algorithms for Complex Optimisation Problems. Mathematical optimisation gives a systematic way for optimal decision-making. This project aims to develop new mathematical tools for complex optimisation problems where limited problem information is available. It will generate new foundational theories for alternative optimisation tools, introducing substantial new capability and rigour to the discipline. The project will create significant new mathematical optimisation techn ....Robust Derivative-Free Algorithms for Complex Optimisation Problems. Mathematical optimisation gives a systematic way for optimal decision-making. This project aims to develop new mathematical tools for complex optimisation problems where limited problem information is available. It will generate new foundational theories for alternative optimisation tools, introducing substantial new capability and rigour to the discipline. The project will create significant new mathematical optimisation techniques and create world-leading and publicly available software. These new techniques and software may ultimately be able to solve some of the most complex optimisation problems in research and industry, such as improving long-term climate predictions and designing 3D-printed medical implants.Read moreRead less
Efficient Computational Methods for Constrained Path Problems. We consider a class of path design problems which arise when an object needs to traverse between two points through a specified region. The region may be a continuous space or the path may be restricted to the edges of a network. The path must optimise a prescribed criterion such
as risk, reliability or cost and satisfy a number of constraints.
Problems of this type readily arise in the defence, transport and
communication i ....Efficient Computational Methods for Constrained Path Problems. We consider a class of path design problems which arise when an object needs to traverse between two points through a specified region. The region may be a continuous space or the path may be restricted to the edges of a network. The path must optimise a prescribed criterion such
as risk, reliability or cost and satisfy a number of constraints.
Problems of this type readily arise in the defence, transport and
communication industries. In addition to efficient solution methods
for these problems the project will produce computational tools for
a wide range of related network routing problems.Read moreRead less
Using global optimization technique to determine the most efficient use of building/floor space to accommodate a given office design. The commercial property market is one of the largest business markets, both in Australia and globally. Businesses of all kinds use commercial office space, which represents many billions of investment dollars. A better understanding of what constitutes efficient and effective office space would produce enormous commercial benefits for this country. Historically, ....Using global optimization technique to determine the most efficient use of building/floor space to accommodate a given office design. The commercial property market is one of the largest business markets, both in Australia and globally. Businesses of all kinds use commercial office space, which represents many billions of investment dollars. A better understanding of what constitutes efficient and effective office space would produce enormous commercial benefits for this country. Historically, very little (if any) consideration has been given to the efficiency of office space design. The measurement of efficiency has now become an essential component of 'site selection'. To date, the application of optimization methodologies have not been applied to the architectural industry, making the development of tools to address this problem a significant and innovative move.Read moreRead less
Economic Scheduling for Efficient Management of Clusters and their Cooperative Federation. Clusters of commodity computers have emerged as mainstream parallel and distributed platforms for high-performance computing. They are presented together as a single, unified resource to the end users by middleware technologies such as resource management and scheduling (RMS) systems. However, existing cluster RMS systems continue to use system centric models rather than utility models for the management a ....Economic Scheduling for Efficient Management of Clusters and their Cooperative Federation. Clusters of commodity computers have emerged as mainstream parallel and distributed platforms for high-performance computing. They are presented together as a single, unified resource to the end users by middleware technologies such as resource management and scheduling (RMS) systems. However, existing cluster RMS systems continue to use system centric models rather than utility models for the management and allocation of resources. There is also little emphasis on the construction of a cooperative federation of clusters to facilitate transparent sharing of resources. To enhance the value delivered by shared clusters, we propose the use of computational economy metaphor in resource management. This project aims to develop (A) computational economy based scheduling policies for allocation of resources and (B) a software infrastructure for creation of cooperative federation of distributed clusters.Read moreRead less
Effective computational methods for nonlinear cone optimisation with industrial applications. This project brings together a number of national and international researchers whose combined expertise will focus on solving optimisation problems arising in a range of industries. The work will result in new cutting edge optimisation technology that can benefit industry and the community.
Discovery Early Career Researcher Award - Grant ID: DE150100240
Funder
Australian Research Council
Funding Amount
$315,000.00
Summary
Geometry and Conditioning in Structured Conic Problems. Conic programming allows one to model and solve large industrial problems via modern optimisation methods, such as interior-point algorithms. These methods are efficient and reliable in solving a vast number of problems, however, they fail on a relatively small but significant set of ill-posed instances, thus affecting the overall reliability of the technique. The reason for such behaviour is profound and constitutes one of the major unsolv ....Geometry and Conditioning in Structured Conic Problems. Conic programming allows one to model and solve large industrial problems via modern optimisation methods, such as interior-point algorithms. These methods are efficient and reliable in solving a vast number of problems, however, they fail on a relatively small but significant set of ill-posed instances, thus affecting the overall reliability of the technique. The reason for such behaviour is profound and constitutes one of the major unsolved problems in real complexity: there is no known algorithm that solves conic problems with real data in polynomial time. The project aims to develop a deep understanding of the geometry of conic problems, aiming for the resolution of this fundamental problem in computational theory.Read moreRead less
Realising the promise of neural networks for practical optimisation: improving their efficiency and effectivess through chaotic dynamics and hardware implementation. Combinatorial optimisation problems such as transportation routing and assembly-line scheduling are critical to the efficiency of many industries, but their combinatorial explosion makes rapid solution difficult. Neural networks (NNs) hold much potential for rapid solution though hardware implementation, but we need to improve the q ....Realising the promise of neural networks for practical optimisation: improving their efficiency and effectivess through chaotic dynamics and hardware implementation. Combinatorial optimisation problems such as transportation routing and assembly-line scheduling are critical to the efficiency of many industries, but their combinatorial explosion makes rapid solution difficult. Neural networks (NNs) hold much potential for rapid solution though hardware implementation, but we need to improve the quality of their solutions before developing hardware. We have previously shown that the rich dynamics of chaos can improve the efficiency and effectiveness of NNs. We aim to develop new chaotic NN models, rigorously evaluate them on industrially significant problems such as those arising in manufacturing, logistics and telecommunications, and demonstrate their speed through hardware acceleration.Read moreRead less
Unlocking the potential for linear and discrete optimisation in knot theory and computational topology. Computational topology is a young, energetic field that uses computers to solve complex geometric problems, such as whether a loop of string is tangled. Such computations are becoming increasingly important in mathematics, and applications span biology, physics and information sciences, however many core problems in the field remain intractable for all but the simplest cases. This project unit ....Unlocking the potential for linear and discrete optimisation in knot theory and computational topology. Computational topology is a young, energetic field that uses computers to solve complex geometric problems, such as whether a loop of string is tangled. Such computations are becoming increasingly important in mathematics, and applications span biology, physics and information sciences, however many core problems in the field remain intractable for all but the simplest cases. This project unites geometric techniques with powerful methods from operations research, such as linear and discrete optimisation, to build fast, powerful tools that can for the first time systematically solve large topological problems. Theoretically, this project has significant impact on the famous open problem of detecting knottedness in fast polynomial time.Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE180100923
Funder
Australian Research Council
Funding Amount
$348,575.00
Summary
Efficient second-order optimisation algorithms for learning from big data. This project aims to apply a diverse range of scientific computing techniques to design and implement new, second-order methods that can surpass first-order alternatives in the next generation of optimisation methods for large-scale machine learning (ML). Scalable optimisation methods are now an integral part ML in the presence of “big data”. While the development of efficient first-order methods has grown in the ML comm ....Efficient second-order optimisation algorithms for learning from big data. This project aims to apply a diverse range of scientific computing techniques to design and implement new, second-order methods that can surpass first-order alternatives in the next generation of optimisation methods for large-scale machine learning (ML). Scalable optimisation methods are now an integral part ML in the presence of “big data”. While the development of efficient first-order methods has grown in the ML community, second-order alternatives have largely been ignored. The project expects to facilitate the development of more effective ML algorithms for extraction of knowledge from large data sets.Read moreRead less
Algorithms for hard graph problems based on auxiliary data. When solving computational problems, algorithms usually access only the data that is absolutely necessary to define the problem. However, much more data is often readily available. Especially for important or slowly evolving data, such as road networks, social graphs, company rankings, or molecules, more and more auxiliary data becomes available through computational processes, sensors, and simple user entries. This auxiliary data can g ....Algorithms for hard graph problems based on auxiliary data. When solving computational problems, algorithms usually access only the data that is absolutely necessary to define the problem. However, much more data is often readily available. Especially for important or slowly evolving data, such as road networks, social graphs, company rankings, or molecules, more and more auxiliary data becomes available through computational processes, sensors, and simple user entries. This auxiliary data can greatly speed up an algorithm and improve its accuracy. This project aims to design improved algorithms that harness auxiliary data to solve selected high-impact NP-hard graph problems, and will build a new empowering theory to discern when auxiliary data can be used to improve algorithms.Read moreRead less