Predicting extremes when events occur in bursts

预测事件突发时的极端情况

• 批准号: DE160101147
• 负责人: Dr Peter Straka
• 金额: $ 20.62万
• 依托单位: The University of New South Wales
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Early Career Researcher Award
• 财政年份: 2016
• 资助国家: 澳大利亚
• 起止时间: 2016-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DE160101147
• 关键词: Predicting; extremes; when; events; occur

This project seeks to advance knowledge in extreme value theory. Extreme value theory is essential to quantify risks in complex systems, such as the risk of network failures. Current statistical models for the occurrence of extremes assume that events happen regularly. This assumption, however, is at odds with human actions and many biological and physical events, which occur in bursts. There is a strong need to understand the effect of such ‘bursty dynamics’ on the frequency and magnitude of extreme events. This project aims to develop extreme value theory for bursty events and thus lay the mathematical groundwork for the estimation and prediction of extremes in a variety of scientific contexts.

这个项目旨在推进极值理论的知识。极值理论对于量化复杂系统中的风险至关重要，例如网络故障风险。目前关于极端事件发生的统计模型假设事件有规律地发生。然而，这种假设与人类活动和许多生物和物理事件不一致，这些事件都是突然发生的。我们迫切需要了解这种“突发动力学”对极端事件发生频率和强度的影响。本项目旨在发展突发事件的极值理论，从而为各种科学背景下的极值估计和预测奠定数学基础。