A5081147777- Statistical Methods and Inference
- Advanced Statistical Methods and Models
- Statistical and numerical algorithms
- Health disparities and outcomes
- Control Systems and Identification
- Image and Signal Denoising Methods
- Statistical Methods and Bayesian Inference
- Optimal Experimental Design Methods
- Evolutionary Algorithms and Applications
- Advanced Multi-Objective Optimization Algorithms
- Healthcare Policy and Management
- Metaheuristic Optimization Algorithms Research
- Data-Driven Disease Surveillance
- Probabilistic and Robust Engineering Design
- Breastfeeding Practices and Influences
- Sparse and Compressive Sensing Techniques
- Neural Networks and Applications
- Gaussian Processes and Bayesian Inference
- Dementia and Cognitive Impairment Research
- Medical Coding and Health Information
- Forecasting Techniques and Applications
- Forest ecology and management
- Forensic Anthropology and Bioarchaeology Studies
- Health Systems, Economic Evaluations, Quality of Life
- Financial Risk and Volatility Modeling
The University of Western Australia
2015-2025
Murdoch University
2021
National University of Singapore
2008-2009
Australian National University
1995-2000
The University of Adelaide
1999
UCLouvain
1992-1995
The title Lasso has been suggested by Tibshirani (1996) as a colourful name for technique of variable selection which requires the minimization sum squares subject to an l1 bound κ on solution. This forces zero components in minimizing solution small values κ. Thus this can function parameter. paper makes two contributions computational problems associated with implementing Lasso: (1) compact descent method solving constrained problem particular value is formulated, and (2) homotopy method,...
Abstract Proposed by Tibshirani, the least absolute shrinkage and selection operator (LASSO) estimates a vector of regression coefficients minimizing residual sum squares subject to constraint on l 1-norm coefficient vector. The LASSO estimator typically has one or more zero elements thus shares characteristics both estimation variable selection. In this article we treat as convex programming problem derive its dual. Consideration primal dual problems together leads important new insights...
We propose a new method for selecting common subset of explanatory variables where the aim is to model several response variables. The idea natural extension LASSO technique proposed by Tibshirani (1996) and based on (joint) residual sum squares while constraining parameter estimates lie within suitable polyhedral region. properties resulting convex programming problem are analyzed special case an orthonormal design. For general case, we develop efficient interior point algorithm....
Abstract Proposed by Tibshirani, the least absolute shrinkage and selection operator (LASSO) estimates a vector of regression coefficients minimizing residual sum squares subject to constraint on l 1-norm coefficient vector. The LASSO estimator typically has one or more zero elements thus shares characteristics both estimation variable selection. In this article we treat as convex programming problem derive its dual. Consideration primal dual problems together leads important new insights...
Human milk is a complex and variable ecosystem fundamental to the development of newborns. This study aimed investigate relationships between human oligosaccharides (HMO) bacterial profiles infant body composition. samples (n = 60) were collected at two months postpartum. Infant maternal composition was measured with bioimpedance spectroscopy. assessed using full-length 16S rRNA gene sequencing 19 HMOs quantitated high-performance liquid chromatography. Relative abundance taxa significantly...
There are a wide array of smoothing methods available for finding structure in data. A general framework is developed which shows that many these can be viewed as projection the data, with respect to appropriate norms. The underlying vector space an unusually large product space, allows inclusion range smoothers our setup (including not typically considered projections). We give several applications this simple geometric interpretation smoothing. major payoff natural and computationally...
Genetic algorithms have been extensively used and studied in computer science, yet there is no generally accepted methodology for exploring which parameters significantly affect performance, whether any interaction between parameters, how performance varies with respect to changes parameters. This paper presents a rigorous practical statistical the exploratory study of genetic other adaptive algorithms. addresses issues experimental design, blocking, power calculations, response curve...
All analyses of spatially aggregated data are vulnerable to the modifiable areal unit problem (MAUP), which describes sensitivity analytical results arbitrary choice spatial aggregation at measured. The MAUP is a serious endemic in all scientific disciplines. However, impact rarely considered, perhaps partly because it still widely considered be unsolvable. It was originally understood that solution should constitute comprehensive statistical framework describing regularities estimates...
We introduce interpolation methods that enable nonlinear wavelet estimators to be employed with stochastic design, or nondyadic regular in problems of nonparametric regression. This approach allows relatively rapid computation, involving dyadic approximations wavelet-after-interpolation techniques. New types are described, enabling first-order variance reduction at the expense second-order increases bias. The effect on threshold choice is addressed, and appropriate thresholds suggested for...
Temporal development of maternal and infant microbiomes during early life impacts short- long-term health. This study aimed to characterize bacterial dynamics within faecal, human milk (HM), oral, faecal samples the exclusive breastfeeding period document associations between oligosaccharide (HMO) intakes oral profiles. Maternal (n = 10) were collected at 2−5, 30, 60, 90 120 days postpartum full-length 16S ribosomal RNA (rRNA) gene was sequenced. Nineteen HMOs quantitated using...
Abstract Background In disease mapping, fine-resolution spatial health data are routinely aggregated for various reasons, example to protect privacy. Usually, such aggregation occurs only once, resulting in ‘single-aggregation maps’ whose representation of the underlying depends on chosen set units. This dependence is described by modifiable areal unit problem (MAUP). Despite an extensive literature, practice, MAUP rarely acknowledged, including mapping. Further, despite single-aggregation...
This book describes an interactive statistical computing environment called 1 XploRe. As the name suggests, support for exploratory analysis is given by a variety of computational tools. XploRe matrix-oriented language with comprehensive set basic operations that provides highly graphics, as well programming environ ment user-written macros; it offers hard-wired smoothing procedures effective high-dimensional data analysis. Its dynamic graphic capa bilities make possible to construct...
We describe a study of the discrimination early melanoma from common and dysplastic nevus using fiber optic diffuse reflectance spectroscopy. Diffuse spectra in wavelength range 550 to 1000 nm are obtained 400-µm core multimode fibers arranged six-illumination-around-one-collection geometry with single fiber-fiber spacing 470 µm. Spectra collected at specific locations on 120 pigmented lesions selected by clinicians as possible melanoma, including 64 histopathologically diagnosed melanoma....
ABSTRACT Objectives To identify changes in orthodontic management strategies patients with hypodontia seen 2000, 2010, and 2017/2018 (during a 1-year period). Materials Methods An assessment of the panoramic radiographs 3701 from Western Australian private practice identified 276 individuals demonstrating hypodontia. The location missing teeth, age, sex, type malocclusion, (space closure or opening) for each patient were noted. Results Most involved agenesis three fewer teeth (90%)....
ABSTRACT End‐range movements are among the most demanding but least understood in sport of tennis. Using male Hawk‐Eye data from match‐play during 2021–2023 Australian Open tournaments, we evaluated speed, deceleration, acceleration, and shot quality characteristics these types movement men's Grand Slam Lateral end‐range that incorporated a change direction (CoD) were identified for analysis using k‐means (end‐range) random forest machine learning models. Peak average deceleration into CoD,...
The aim of this study was to determine the comparative accuracy Demirjian's four dental development methods for forensic age estimation in Western Australian population. A sample comprising 143 individuals aged 4.6 14.5 years were assessed using (original 7-tooth: M(2), M(1), PM(2), PM(1), C, I(2), and I(1); revised 4-tooth: PM(1); an alternate I(1)). When comparing all methods, 4-tooth method overestimated both males females by 0.04 0.25 years, respectively. original 7-tooth least accurate...