Advancing statistical models for clustering data with structured dependence. Modern data present increasingly complexities such as heterogeneity and structured dependence among data. Ignoring these fe

Description

Advancing statistical models for clustering data with structured dependence. Modern data present increasingly complexities such as heterogeneity and structured dependence among data. Ignoring these features can result in misleading findings. This project aims to develop novel methods to identify important subgroups in data with various forms of dependence. It will introduce techniques that can capture complex relationships in data and enhance model validity. Main outcomes include advanced methods and algorithms that can accurately identify clusters, patterns, outliers, and model evaluation. This will provide significant benefits in statistics and for crime prevention in Australia when the new methods are applied to Queensland Police Service data to understand co-offending crimes, repeat victimisation, and hot spots.. Scheme: Discovery Projects. Field: 4905 - Statistics. Lead: Prof Shu-Kay Angus Ng