A5068071581- Statistical Methods and Bayesian Inference
- Bayesian Methods and Mixture Models
- Statistical Methods and Inference
- Reliability and Maintenance Optimization
- Geological Modeling and Analysis
- Soil Geostatistics and Mapping
- Probabilistic and Robust Engineering Design
- Wind and Air Flow Studies
- Reservoir Engineering and Simulation Methods
- Fault Detection and Control Systems
- Statistical Distribution Estimation and Applications
- Seismic Imaging and Inversion Techniques
- Financial Risk and Volatility Modeling
- Coral and Marine Ecosystems Studies
- Structural Health Monitoring Techniques
- Geophysical and Geoelectrical Methods
- Spectroscopy and Chemometric Analyses
- Geochemistry and Geologic Mapping
- Spatial and Panel Data Analysis
- Marine and fisheries research
- Model Reduction and Neural Networks
- Advanced Statistical Methods and Models
- Oceanographic and Atmospheric Processes
- Marine and coastal plant biology
- Structural Integrity and Reliability Analysis
The University of Western Australia
2015-2024
ARC Centre of Excellence in Advanced Molecular Imaging
2022
The University of Sydney
2019
Australian Institute of Marine Science
2010
UNSW Sydney
2003-2006
University of Missouri
2005
Significance Tests of biodiversity theory have been controversial partly because alternative formulations the same seemingly yield different conclusions. This has a particular challenge for neutral theory, which dominated tests over last decade. Neutral attributes differences in species abundances to chance variation individuals’ fates, rather than traits. By identifying common features models, we conduct uniquely robust test across global dataset marine assemblages. Consistently, vary more...
Abstract. Unlike some other well-known challenges such as facial recognition, where machine learning and inversion algorithms are widely developed, the geosciences suffer from a lack of large, labelled data sets that can be used to validate or train robust schemes. Publicly available 3D geological models far too restricted in both number range scenarios serve these purposes. With reference inverting geophysical this problem is further exacerbated most cases real observations result unknown...
The Digital Twin (DT) paradigm offers an extension of simulation model utility into the operational phase engineering asset. goal is a "twinned" with observed data that reflects actual performance However, exploring sources uncertainty for both physical asset and are challenge. For example, random metocean conditions, on parameters response behaviour offshore wind turbine (OWT) structures, contribute to predicted life under fatigue. In-service assessment OWT structures will benefit from...
Abstract. Parametric geological models such as implicit or kinematic provide low-dimensional, interpretable representations of 3-D structures. Combining these with geophysical data in a probabilistic joint inversion framework provides an opportunity to directly quantify uncertainty interpretations. For best results, care must be taken the intermediate step rendering parametric geology finite-resolution discrete basis for calculation. Calculating geophysics from naively voxelized geology,...
The past two decades have seen a rapid adoption of artificial intelligence methods applied to mineral exploration. More recently, the easier acquisition some types data has inspired broad literature that examined many machine learning and modelling techniques combine exploration criteria, or 'features', generate predictions for prospectivity. Central design prospectivity models is 'mineral system', conceptual model describing key geological elements control timing location economic...
Organizations are increasingly making use of communities practice (CoPs) as a way leveraging the dispersed knowledge and expertise their employees. One important in which CoPs predicted to benefit organizations is by facilitating transfer best practices. In this study, we examined impact introduction global on changes made operational procedures three refineries operated multinational company over period more than 5 years. We used Bayesian change point detection model assess probability that...
Significance Science and engineering have benefited greatly from the ability of finite element methods (FEMs) to simulate nonlinear, time-dependent complex systems. The recent advent extensive data collection such systems now raises question how systematically incorporate these into models, consistently updating solution in face mathematical model misspecification with physical reality. This article describes general widely applicable methodology for coherent synthesis FEM providing a...
Statistical learning additions to physically derived mathematical models are gaining traction in the literature. A recent approach has been augment underlying physics of governing equations with data driven Bayesian statistical methodology. Coined statFEM, method acknowledges a priori model misspecification, by embedding stochastic forcing within equations. Upon receipt additional data, posterior distribution discretised finite element solution is updated using classical filtering...
A load-sharing system has several components and the failure of one component can affect lifetime surviving components. Since does not equate to for different designs, analysis dependence structure between becomes a meaningful exercise. The extended sequential order statistics model allows us heterogeneous in systems. However, results may suggest that risk decreases as fail sequentially, which be counterintuitive, especially when data are scarce. We propose address this issue by imposing an...
Abstract The model of extended sequential order statistics (ESOS) comprises two valuable characteristics making the powerful when modelling multi‐component systems. First, components can be assumed to heterogeneous and second, component lifetime distributions change upon failure other components. This degree flexibility comes at cost a large number parameters. exact depends on system size observation depth quickly exceed observations available. Consequently, would benefit from reduction in...
Abstract Welch’s method provides an estimator of the power spectral density that is statistically consistent. This achieved by averaging over periodograms calculated from overlapping segments a time series. For finite-length series, while variance decreases as number increases, magnitude estimator’s bias increases: bias-variance trade-off ensues when setting segment number. We address this issue providing novel for debiasing maintains computational complexity and asymptotic consistency,...
The time-dependent behavior in the variability of wind measurements is investigated using bivariate generalized autoregressive conditional heteroscedastic models. These models express current level short-timescale terms previous observed values fluctuations from mean fields. As such, these provide a useful descriptive model that can be applied to short-term forecasting around local levels.
Land cover data derived from satellites are commonly used to prescribe inputs models of the land surface. Since such inevitably contains errors, quantifying how uncertainties in affect a model’s output is important. To do so, spatial distribution possible values required propagate through simulation. However, at large scales, as those for climate models, modelling can be difficult. Also, computer often require proportions sites larger than original map scale inputs, and it uncertainty these...
Journal Article Bayesian estimation of a random effects heteroscedastic probit model Get access Yuanyuan Gu, Gu School Economics, University New South Wales, Sydney, NSW 2052, Australia Search for other works by this author on: Oxford Academic Google Scholar Denzil G. Fiebig, Fiebig Edward Cripps, Cripps Mathematics and Statistics, Western Australia, Crawley, WA 6009, Robert Kohn The Econometrics Journal, Volume 12, Issue 2, 1 July 2009, Pages 324–339,...
Abstract In this article, we will address the complexity of non‐identical components in multi‐component systems. Most technical systems can be described as such since either component types or functions within system vary amongst components. While most reliability related work resorts to assumption homogeneous components, aim often more realistic heterogeneous extending model Extended Sequential Order Statistics by two inferential methods. Firstly, derivation Maximum Likelihood Estimates...