Novel statistical methods for data with non-Euclidean geometric structure. This project aims to develop new flexible regression models and classification algorithms, along with robust and efficient inference methods, applicable to a wide range of non-Euclidean data types which arise in many fields of science, business and technology. There are serious flaws with currently available methods of analysis for non-Euclidean data. This project expects to transform such analyses by providing new quanti ....Novel statistical methods for data with non-Euclidean geometric structure. This project aims to develop new flexible regression models and classification algorithms, along with robust and efficient inference methods, applicable to a wide range of non-Euclidean data types which arise in many fields of science, business and technology. There are serious flaws with currently available methods of analysis for non-Euclidean data. This project expects to transform such analyses by providing new quantitative tools within a unifying framework. The anticipated project outcomes will be of mathematical interest and valuable in applications such as finance (predicting Australian stock returns); modelling electroencephalography data; Australian geochemical data, relating to sediments; and Australian X-ray tumour image data. Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE180100220
Funder
Australian Research Council
Funding Amount
$369,075.00
Summary
Statistics for manifold-valued data. This project aims to develop, and then implement, a new suite of fully flexible, interpretable and tractable models for manifold-valued data, along with robust and accurate estimation techniques for their parameters. Multivariate data with complicated constraints, such as manifold-valued data, is frequently encountered in the physical, biological and medical sciences, however it is difficult to define tractable statistical models and estimate their parameters ....Statistics for manifold-valued data. This project aims to develop, and then implement, a new suite of fully flexible, interpretable and tractable models for manifold-valued data, along with robust and accurate estimation techniques for their parameters. Multivariate data with complicated constraints, such as manifold-valued data, is frequently encountered in the physical, biological and medical sciences, however it is difficult to define tractable statistical models and estimate their parameters due to the curvature and nonlinear geometry of the sample space. The outcomes of the project are of direct mathematical interest as well as having significant interest to science and business disciplines where manifold-valued data is commonly observed.Read moreRead less
New statistical tools for mineral exploration targeting and validation. Exploration for new mineral resources depends on information gleaned from geological survey data. This project confronts important, unsolved statistical problems in the analysis of geological survey data which have direct impact on exploration targeting.
Discovery Early Career Researcher Award - Grant ID: DE180100203
Funder
Australian Research Council
Funding Amount
$348,575.00
Summary
Deep space-time models for modelling complex environmental phenomena. This project aims to adapt deep-learning models, used in areas of artificial intelligence such as image tagging and automatic text translation, to improve our understanding of the environment. The project expects to develop new theory for deep-learning models to learn from measurement data and numerical-model output about environmental phenomena that evolve in space and time, such as ice sheets and the atmosphere. Expected out ....Deep space-time models for modelling complex environmental phenomena. This project aims to adapt deep-learning models, used in areas of artificial intelligence such as image tagging and automatic text translation, to improve our understanding of the environment. The project expects to develop new theory for deep-learning models to learn from measurement data and numerical-model output about environmental phenomena that evolve in space and time, such as ice sheets and the atmosphere. Expected outcomes include the ability to provide reliable predictions and quantification of uncertainty on environmental concerns of national importance, such as sea-level rise. Key benefits include improved risk management and mitigation, for example through financial incentives or infrastructure planning.Read moreRead less
New Directions in Bayesian Statistics: formulation, computation and application to exemplar challenges. Bayesian statistics is a fundamental statistical and machine learning approach for density estimation, data analysis and inference. However, there remain open questions regarding the formulation of the model, the likelihood and priors, and efficient computation. This project proposes new approaches that address these issues, and applies them to two exemplar challenges: the impact of climate ch ....New Directions in Bayesian Statistics: formulation, computation and application to exemplar challenges. Bayesian statistics is a fundamental statistical and machine learning approach for density estimation, data analysis and inference. However, there remain open questions regarding the formulation of the model, the likelihood and priors, and efficient computation. This project proposes new approaches that address these issues, and applies them to two exemplar challenges: the impact of climate change on the Great Barrier Reef and better understanding neurological diseases related aging, in particular Parkinson's Disease. Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE200100200
Funder
Australian Research Council
Funding Amount
$418,398.00
Summary
Next generation causal inference methods for biological data. This project aims to develop next generation causal inference methods for analysing biological data especially the single cell sequencing data and their applications in cell biology. Although Artificial Intelligence and Statistical Machine Learning have been applied successfully in many fields, including biological research, there is still a serious lack of methods for interpreting and reasoning about the mechanism of biological syste ....Next generation causal inference methods for biological data. This project aims to develop next generation causal inference methods for analysing biological data especially the single cell sequencing data and their applications in cell biology. Although Artificial Intelligence and Statistical Machine Learning have been applied successfully in many fields, including biological research, there is still a serious lack of methods for interpreting and reasoning about the mechanism of biological systems, the ultimate goal of research in many areas. Efficient data-driven causality discovery approaches developed by the project will be a timely and significant contribution to the knowledge of biology and statistics as well as the battle against health threats.
Read moreRead less
Generalised Degrees of Freedom and Probabilistic Regularisation. This project intends to develop novel statistical tools for more accurate prediction by taking account of model complexity and uncertainties associated with the fitting procedure. The project also plans to develop a novel shrinkage approach via new penalty functions to avoid over-fitting and asymptotic properties. The key applications may include genetic studies where the number of predictors is large and biological experiments whe ....Generalised Degrees of Freedom and Probabilistic Regularisation. This project intends to develop novel statistical tools for more accurate prediction by taking account of model complexity and uncertainties associated with the fitting procedure. The project also plans to develop a novel shrinkage approach via new penalty functions to avoid over-fitting and asymptotic properties. The key applications may include genetic studies where the number of predictors is large and biological experiments where multivariate and temporal data are often collected – for example economical breeding in animal and fish farming and more effectively detecting the genes of interest in genetic studies on human, animals and plants.Read moreRead less
Robust inferences for analysis of longitudinal data. This project will develop novel statistical tools. Outcomes of this project will enable more reliable data analysis and more cost effective designs in environmental and biological studies.
Precision ecology: the modern era of designed experiments in plant ecology. This project aims to develop the field of precision ecology, forging a new era of designed experiments where sampling is informed by research questions and what is known about the ecological process being studied. Through the development of novel statistical methods, new experiments globally will be designed to answer important ecological questions including what influence abiotic and biotic factors have on plant commun ....Precision ecology: the modern era of designed experiments in plant ecology. This project aims to develop the field of precision ecology, forging a new era of designed experiments where sampling is informed by research questions and what is known about the ecological process being studied. Through the development of novel statistical methods, new experiments globally will be designed to answer important ecological questions including what influence abiotic and biotic factors have on plant communities over time and different spatial scales. Expected outcomes include new methods and tools that will modernise how future experiments will be conducted in plant ecology. This will provide significant transdisciplinary benefits including new statistical methods that target scientific discovery in ecological studies.Read moreRead less
Principled statistical methods for high-dimensional correlation networks. This project aims to develop a novel and principled approach for building correlation networks. Correlation networks aim to identify the most significant associations present in modern massive datasets, and have numerous applications, ranging from the biomedical and environmental sciences to the social sciences. Nodes of such networks represent features, and edges represent associations, or the lack thereof. Current method ....Principled statistical methods for high-dimensional correlation networks. This project aims to develop a novel and principled approach for building correlation networks. Correlation networks aim to identify the most significant associations present in modern massive datasets, and have numerous applications, ranging from the biomedical and environmental sciences to the social sciences. Nodes of such networks represent features, and edges represent associations, or the lack thereof. Current methods are not readily scalable to modern ultra-high dimensional settings, and do not account for uncertainty in the estimated associations. This project will develop a principled, highly scalable methodology for building such networks, which incorporates uncertainty quantification. Emphasis is placed on modern ultra-high dimensional settings in which differentiating a true correlation from a spurious one is a notoriously difficult task.Read moreRead less