arXiv subjects

Kerrie Mengersen

Publications and source records attributed to Kerrie Mengersen.

At least 19 recordsLinked to original sources

arXiv2026

A Sparse-Group Pliable Lasso

The sparse-group pliable Lasso (SGPL) extends the pliable Lasso and group pliable Lasso by combining sparse-group regularization with a predictor-level coupling penalty, enabling simultaneous group-level selection, within-group sparsity, and hierarchical structure between main effects and interactions. We propose a blockwise coordinate descent algorithm for fitting the SGPL that exploits the convexity and structure of the objective function, establish convexity and Karush--Kuhn--Tucker optimality conditions, and prove that the algorithm converges to a global minimizer. Simulation studies demonstrate competitive predictive performance and smaller interaction estimation error than the pliable Lasso and group pliable Lasso, albeit with the expected precision--recall trade-off in support recovery. We further illustrate the proposed method using a Parkinson's disease gut microbiome study and an adrenocortical carcinoma (ACC) copy-number dataset from The Cancer Genome Atlas. The Parkinson's application identifies interpretable interactions between microbial abundances and dietary variables, while the ACC application illustrates that the effectiveness of group-structured regularization depends on how well the prespecified grouping reflects the underlying signal structure.

stat.ME↗

arXiv2026

Hierarchical Bayes meets hierarchical forecasting: A flexible framework for level-focused forecasts

Decision-making in hierarchical systems requires probabilistic forecasts at all cross-sectional levels. Current hierarchical forecasting methods typically generate independent forecasts at each level and reconcile them post hoc to ensure coherence between upper and lower levels. Such post hoc corrections do not incorporate hierarchical structure or decision goals into the underlying parameter estimation. We propose a fully Bayesian hierarchical forecasting framework that shares information more effectively between and across levels than reconciliation alone. Our approach has the flexibility to softly penalise incoherence, subject to model specification, and to focus the global model and coherence update on hierarchical levels most relevant to decision outcomes. This yields parameter estimates that are focused towards the forecasting goals and capture the requirement for coherency, removing the need to estimate covariance matrices for multi-step forecasting horizons. We demonstrate improvements in predictive accuracy metrics on both simulated data and Australian domestic tourism forecasting.

stat.ME↗

arXiv2026

Piecewise Deterministic Markov Processes for Bayesian Neural Networks

Inference on modern Bayesian Neural Networks (BNNs) often relies on a variational inference treatment, imposing violated assumptions of independence and the form of the posterior. Traditional MCMC approaches avoid these assumptions at the cost of increased computation due to its incompatibility to subsampling of the likelihood. New Piecewise Deterministic Markov Process (PDMP) samplers permit subsampling, though introduce a model specific inhomogenous Poisson Process (IPPs) which is difficult to sample from. This work introduces a new generic and adaptive thinning scheme for sampling from these IPPs, and demonstrates how this approach can accelerate the application of PDMPs for inference in BNNs. Experimentation illustrates how inference with these methods is computationally feasible, can improve predictive accuracy, MCMC mixing performance, and provide informative uncertainty measurements when compared against other approximate inference schemes.

stat.ML↗

arXiv2026

Bayesian spatiotemporal modelling of political violence and conflict events using discrete-time Hawkes processes

The monitoring of conflict risk in the humanitarian sector is largely based on simple historic averages. The overarching goal of this work is to assess the potential for using a more statistically rigorous approach to monitor the risk of political violence and conflict events in practice, and thereby improve our understanding of their temporal and spatial patterns, to inform preventative measures. In particular, a Bayesian, spatiotemporal variant of the Hawkes process is fitted to data gathered by the Armed Conflict Location and Event Data (ACLED) project to obtain sub-national estimates of conflict risk in South Asia over time and space. Our model can effectively estimate the risk level of these events within a statistically sound framework, with a more precise understanding of uncertainty than was previously possible. The model also provides insights into differences in behaviours between countries and conflict types. We also show how our model can be used to monitor short and long term trends, and that it is more stable and robust to outliers compared to current practices that rely on historical averages.

stat.AP↗

arXiv2025

Optimal Clustering with Dependent Costs in Bayesian Networks

Background: Clustering of nodes in Bayesian Networks (BNs) and related graphical models such as Dynamic BNs (DBNs) has been demonstrated to enhance computational efficiency and improve model learning. It typically involves partitioning the underlying Directed Acyclic Graph (DAG) into cliques or optimising for some cost or criteria. Objectives: We focus on a critical but understudied aspect of optimal clustering involving cost dependency. This is where inference outcomes and hence clustering costs depend on both nodes within a cluster and the mapping of clusters that are connected by at least one arc. Methods: We propose a novel algorithm called Dependent Cluster MAPping (DCMAP) which can, given an arbitrary, positive cost function, iteratively and rapidly find near-optimal, then optimal cluster mappings. Results: DCMAP is shown analytically to be optimal in terms of finding all of the least cost cluster mapping solutions and with no more iterations than an equally informed algorithm. Demonstrated on a complex systems seagrass DBN with $9.91\times10^9$ and $1.51\times10^{21}$ possible cluster mappings for 25 and 50 node configurations, it took 856 and 1569 iterations on average to find the first optimal solution, respectively. Conclusions: The effectiveness of DCMAP enables future research in BN learning using optimisation, such as through enhancing computational efficiency or minimising entropy for learning. This is critically important as computation of marginal distributions or updating model parameters is NP-hard.

cs.DS↗

arXiv2025

A Principled Approach to Bayesian Transfer Learning

Updating $\textit{a priori}$ information given some observed data is the core tenet of Bayesian inference. Bayesian transfer learning extends this idea by incorporating information from a related dataset to improve the inference on the observed target dataset which may have been collected under slightly different settings. The use of related information can be useful when the target dataset is scarce, for example. There exist various Bayesian transfer learning methods that decide how to incorporate the related data in different ways. Unfortunately, there is no principled approach for comparing Bayesian transfer methods in real data settings. Additionally, some Bayesian transfer learning methods, such as the so-called power prior approaches, rely on conjugacy or costly specialised techniques. In this paper, we find an effective approach to compare Bayesian transfer learning methods is to apply leave-one-out cross validation on the target dataset. Further, we introduce a new framework, $\textit{transfer sequential Monte Carlo}$, that efficiently implements power prior methods in an automated fashion. We demonstrate the performance of our proposed methods in two comprehensive simulation studies.

stat.ME↗

arXiv2025

Geographic Characterisation of Children's Health and Wellbeing through Vulnerability Indices Using Principal Component Analysis (PCA) and K-Means Clustering

Using Australian children as a case study, this paper aims to develop vulnerability indices to characterise the health and well-being of children aged 0-5 years old. The indices are used to identify differences in children's health and well-being across geographic regions and identify clusters of regions with similar characteristics. The approach is underpinned by two well-known statistical methods, namely Principal Component Analysis and K-Means Clustering. The identification of these regions with similar vulnerability characteristics can then be used to derive new insights into drivers of children's health and well-being and support improved decision-making for services in Australia.

math.GM↗

arXiv2025

AIMSCheck: Leveraging LLMs for AI-Assisted Review of Modern Slavery Statements Across Jurisdictions

Modern Slavery Acts mandate that corporations disclose their efforts to combat modern slavery, aiming to enhance transparency and strengthen practices for its eradication. However, verifying these statements remains challenging due to their complex, diversified language and the sheer number of statements that must be reviewed. The development of NLP tools to assist in this task is also difficult due to a scarcity of annotated data. Furthermore, as modern slavery transparency legislation has been introduced in several countries, the generalizability of such tools across legal jurisdictions must be studied. To address these challenges, we work with domain experts to make two key contributions. First, we present AIMS.uk and AIMS.ca, newly annotated datasets from the UK and Canada to enable cross-jurisdictional evaluation. Second, we introduce AIMSCheck, an end-to-end framework for compliance validation. AIMSCheck decomposes the compliance assessment task into three levels, enhancing interpretability and practical applicability. Our experiments show that models trained on an Australian dataset generalize well across UK and Canadian jurisdictions, demonstrating the potential for broader application in compliance monitoring. We release the benchmark datasets and AIMSCheck to the public to advance AI-adoption in compliance assessment and drive further research in this field.

cs.CY↗

arXiv2025

AIMS.au: A Dataset for the Analysis of Modern Slavery Countermeasures in Corporate Statements

Despite over a decade of legislative efforts to address modern slavery in the supply chains of large corporations, the effectiveness of government oversight remains hampered by the challenge of scrutinizing thousands of statements annually. While Large Language Models (LLMs) can be considered a well established solution for the automatic analysis and summarization of documents, recognizing concrete modern slavery countermeasures taken by companies and differentiating those from vague claims remains a challenging task. To help evaluate and fine-tune LLMs for the assessment of corporate statements, we introduce a dataset composed of 5,731 modern slavery statements taken from the Australian Modern Slavery Register and annotated at the sentence level. This paper details the construction steps for the dataset that include the careful design of annotation specifications, the selection and preprocessing of statements, and the creation of high-quality annotation subsets for effective model evaluations. To demonstrate our dataset's utility, we propose a machine learning methodology for the detection of sentences relevant to mandatory reporting requirements set by the Australian Modern Slavery Act. We then follow this methodology to benchmark modern language models under zero-shot and supervised learning settings.

cs.CL↗

arXiv2024

Unsupervised anomaly detection in spatio-temporal stream network sensor data

The use of in-situ digital sensors for water quality monitoring is becoming increasingly common worldwide. While these sensors provide near real-time data for science, the data are prone to technical anomalies that can undermine the trustworthiness of the data and the accuracy of statistical inferences, particularly in spatial and temporal analyses. Here we propose a framework for detecting anomalies in sensor data recorded in stream networks, which takes advantage of spatial and temporal autocorrelation to improve detection rates. The proposed framework involves the implementation of effective data imputation to handle missing data, alignment of time-series to address temporal disparities, and the identification of water quality events. We explore the effectiveness of a suite of state-of-the-art statistical methods including posterior predictive distributions, finite mixtures, and Hidden Markov Models (HMM). We showcase the practical implementation of automated anomaly detection in near-real time by employing a Bayesian recursive approach. This demonstration is conducted through a comprehensive simulation study and a practical application to a substantive case study situated in the Herbert River, located in Queensland, Australia, which flows into the Great Barrier Reef. We found that methods such as posterior predictive distributions and HMM produce the best performance in detecting multiple types of anomalies. Utilizing data from multiple sensors deployed relatively near one another enhances the ability to distinguish between water quality events and technical anomalies, thereby significantly improving the accuracy of anomaly detection. Thus, uncertainty and biases in water quality reporting, interpretation, and modelling are reduced, and the effectiveness of subsequent management actions improved.

stat.AP↗

arXiv2024

Mathematical modelling and uncertainty quantification for analysis of biphasic coral reef recovery patterns

Coral reefs are increasingly subjected to major disturbances threatening the health of marine ecosystems. Substantial research underway to develop intervention strategies that assist reefs in recovery from, and resistance to, inevitable future climate and weather extremes. To assess potential benefits of interventions, mechanistic understanding of coral reef recovery and resistance patterns is essential. Recent evidence suggests that more than half of the reefs surveyed across the Great Barrier Reef (GBR) exhibit deviations from standard recovery modelling assumptions when the initial coral cover is low ($\leq 10$\%). New modelling is necessary to account for these observed patterns to better inform management strategies. We consider a new model for reef recovery at the coral cover scale that accounts for biphasic recovery patterns. The model is based on a multispecies Richards' growth model that includes a change point in the recovery patterns. Bayesian inference is applied for uncertainty quantification of key parameters for assessing reef health and recovery patterns. This analysis is applied to benthic survey data from the Australian Institute of Marine Sciences (AIMS). We demonstrate agreement between model predictions and data across every recorded recovery trajectory with at least two years of observations following disturbance events occurring between 1992--2020. This new approach will enable new insights into the biological, ecological and environmental factors that contribute to the duration and severity of biphasic coral recovery patterns across the GBR. These new insights will help to inform managements and monitoring practice to mitigate the impacts of climate change on coral reefs.

stat.AP↗

arXiv2024

Federated Learning for Non-factorizable Models using Deep Generative Prior Approximations

Federated learning (FL) allows for collaborative model training across decentralized clients while preserving privacy by avoiding data sharing. However, current FL methods assume conditional independence between client models, limiting the use of priors that capture dependence, such as Gaussian processes (GPs). We introduce the Structured Independence via deep Generative Model Approximation (SIGMA) prior which enables FL for non-factorizable models across clients, expanding the applicability of FL to fields such as spatial statistics, epidemiology, environmental science, and other domains where modeling dependencies is crucial. The SIGMA prior is a pre-trained deep generative model that approximates the desired prior and induces a specified conditional independence structure in the latent variables, creating an approximate model suitable for FL settings. We demonstrate the SIGMA prior's effectiveness on synthetic data and showcase its utility in a real-world example of FL for spatial data, using a conditional autoregressive prior to model spatial dependence across Australia. Our work enables new FL applications in domains where modeling dependent data is essential for accurate predictions and decision-making.

stat.ML↗

arXiv2024

Scalable Vertical Federated Learning via Data Augmentation and Amortized Inference

Vertical federated learning (VFL) has emerged as a paradigm for collaborative model estimation across multiple clients, each holding a distinct set of covariates. This paper introduces the first comprehensive framework for fitting Bayesian models in the VFL setting. We propose a novel approach that leverages data augmentation techniques to transform VFL problems into a form compatible with existing Bayesian federated learning algorithms. We present an innovative model formulation for specific VFL scenarios where the joint likelihood factorizes into a product of client-specific likelihoods. To mitigate the dimensionality challenge posed by data augmentation, which scales with the number of observations and clients, we develop a factorized amortized variational approximation that achieves scalability independent of the number of observations. We showcase the efficacy of our framework through extensive numerical experiments on logistic regression, multilevel regression, and a novel hierarchical Bayesian split neural net model. Our work paves the way for privacy-preserving, decentralized Bayesian inference in vertically partitioned data scenarios, opening up new avenues for research and applications in various domains.

stat.CO↗

arXiv2024

Guidelines for the Creation of Analysis Ready Data

Globally, there is an increased need for guidelines to produce high-quality data outputs for analysis. No framework currently exists that provides guidelines for a comprehensive approach to producing analysis ready data (ARD). Through critically reviewing and summarising current literature, this paper proposes such guidelines for the creation of ARD. The guidelines proposed in this paper inform ten steps in the generation of ARD: ethics, project documentation, data governance, data management, data storage, data discovery and collection, data cleaning, quality assurance, metadata, and data dictionary. These steps are illustrated through a substantive case study that aimed to create ARD for a digital spatial platform: the Australian Child and Youth Wellbeing Atlas (ACYWA).

cs.DB↗

arXiv2024

Proposer selection in EIP-7251

Immediate settlement, or single-slot finality (SSF), is a long-term goal for Ethereum. The growing active validator set size is placing an increasing computational burden on the network, making SSF more challenging. EIP-7251 aims to reduce the number of validators by giving stakers the option to merge existing validators. Key to the success of this proposal therefore is whether stakers choose to merge their validators once EIP-7251 is implemented. It is natural to assume stakers participate only if they anticipate greater expected utility (risk-adjusted returns) as a single large validator. In this paper, we focus on one of the duties that a validator performs, viz. being the proposer for the next block. This duty can be quite lucrative, but happens infrequently. Based on previous analysis, we may assume that EIP-7251 implies no change to the security of the protocol. We confirm that the probability of a validator being selected as block proposer is equivalent under each consolidation regime. This result ensures that the decision of one staker to merge has no impact on the opportunity of another to propose the next block, in turn ensuring there is no major systemic change to the economics of the protocol with respect to proposer selection.

stat.AP↗

arXiv2024

Creating a Spatial Vulnerability Index for Environmental Health

Extreme natural hazards are increasing in frequency and intensity. These natural changes in our environment, combined with man-made pollution, have substantial economic, social and health impacts globally. The impact of the environment on human health (environmental health) is becoming well understood in international research literature. However, there are significant barriers to understanding key characteristics of this impact, related to substantial data volumes, data access rights and the time required to compile and compare data over regions and time. This study aims to reduce these barriers in Australia by creating an open data repository of national environmental health data and presenting a methodology for the production of health outcome-weighted population vulnerability indices related to extreme heat, extreme cold and air pollution at various temporal and geographical resolutions. Current state-of-the-art methods for the calculation of vulnerability indices include equal weight percentile ranking and the use of principal component analysis (PCA). The weighted vulnerability index methodology proposed in this study offers an advantage over others in the literature by considering health outcomes in the calculation process. The resulting vulnerability percentiles more clearly align population sensitivity and adaptive capacity with health risks. The temporal and spatial resolutions of the indices enable national monitoring on a scale never before seen across Australia. Additionally, we show that a weekly temporal resolution can be used to identify spikes in vulnerability due to changes in relative national environmental exposure.

stat.ME↗

arXiv2024

Spatial Autoregressive Model on a Dirichlet Distribution

Compositional data find broad application across diverse fields due to their efficacy in representing proportions or percentages of various components within a whole. Spatial dependencies often exist in compositional data, particularly when the data represents different land uses or ecological variables. Ignoring the spatial autocorrelations in modelling of compositional data may lead to incorrect estimates of parameters. Hence, it is essential to incorporate spatial information into the statistical analysis of compositional data to obtain accurate and reliable results. However, traditional statistical methods are not directly applicable to compositional data due to the correlation between its observations, which are constrained to lie on a simplex. To address this challenge, the Dirichlet distribution is commonly employed, as its support aligns with the nature of compositional vectors. Specifically, the R package DirichletReg provides a regression model, termed Dirichlet regression, tailored for compositional data. However, this model fails to account for spatial dependencies, thereby restricting its utility in spatial contexts. In this study, we introduce a novel spatial autoregressive Dirichlet regression model for compositional data, adeptly integrating spatial dependencies among observations. We construct a maximum likelihood estimator for a Dirichlet density function augmented with a spatial lag term. We compare this spatial autoregressive model with the same model without spatial lag, where we test both models on synthetic data as well as two real datasets, using different metrics. By considering the spatial relationships among observations, our model provides more accurate and reliable results for the analysis of compositional data. The model is further evaluated against a spatial multinomial regression model for compositional data, and their relative effectiveness is discussed.

stat.ME↗

arXiv2024

Bayesian Design for Sampling Anomalous Spatio-Temporal Data

Data collected from arrays of sensors are essential for informed decision-making in various systems. However, the presence of anomalies can compromise the accuracy and reliability of insights drawn from the collected data or information obtained via statistical analysis. This study aims to develop a robust Bayesian optimal experimental design (BOED) framework with anomaly detection methods for high-quality data collection. We introduce a general framework that involves anomaly generation, detection and error scoring when searching for an optimal design. This method is demonstrated using two comprehensive simulated case studies: the first study uses a spatial dataset, and the second uses a spatio-temporal river network dataset. As a baseline approach, we employed a commonly used prediction-based utility function based on minimising errors. Results illustrate the trade-off between predictive accuracy and anomaly detection performance for our method under various design scenarios. An optimal design robust to anomalies ensures the collection and analysis of more trustworthy data, playing a crucial role in understanding the dynamics of complex systems such as the environment, therefore enabling informed decisions in monitoring, management, and response.

stat.AP↗