Statistical challenges involving indirect data

Grant number: DP170102434 | Funding period: 2018 - 2022

Completed

Abstract

This project aims to develop statistical methodology for solving contemporary problems involving indirectly observed data whose complexity is exacerbated by factors such as incompleteness or episodic availability. Modern statistics find it difficult to analyse complex data which contain important information only in an indirect way, such as data measured with noise or aggregated data. This project considers both finite dimensional data and functional data. The expected methodology will be able to solve frontier problems, where only sophisticated methods can access information. This is expected to benefit brain studies, economics, infectious disease, nutrition and public health.

University of Melbourne Researchers