ANLSDA · Advance Challenging Statistical Analysis for Massive High-Dimensional Nonlinear Spatial Time Series

Modern society, including Europe, faces a variety of global issues, such as global environmental (including climate) changes, global economic and financial crises as well as global energy and sustainable development. Many of these global challenges facing humanity are geo-spatial in nature, with big spatio-temporal data from location-based events or services that are complexly connected and interdependent, requiring advanced, more accurate and effective statistical and econometric techniques in modeling and forecasting of various risks associated with these global challenges.This project aims to develop the cutting-edge methodologies to advance the challenging theoretical and practical issues in statistical inference and econometric analysis of massive high-dimensional nonlinear spatial time series. It will explore some fundamental and difficult issues and establish a unifying novel theory for a framework of non-parametric and semi-parametric approaches to modeling and forecasting of massive nonlinear spatial time series data that involve complex structures and information both from temporal and spatial dimensions arising in such important applications as predicting environmental and climate as well as socioeconomic risks. The developed new generation of statistical and econometric technologies will empower the practitioners and policy-makers to produce more accurate quantitative forecasts that help to generate more informed countermeasures with regard to various risks that our modern society faces.As one of the international pioneering researchers in nonlinear spatial and spatio-temporal data analysis, a new subject in the discipline of Statistics, this project, if supported, will greatly enhance and integrate the principal investigator (PI)’s research and career into the UK and Europe from Australia where the PI received strong financial supports in research from the Australian Research Council.""