摘要：

In recent years, data breaches have become a significant concern, leading to substantial financial losses annually. However, the lack of suitable statistical approaches for assessing breach risks poses an obstacle. To address this challenge, we first propose a novel statistical model that focuses on analyzing hacking breach risks at the individual company level. We then develop a multivariate frequency-severity framework that examines breach risks at the state level. Additionally, we introduce the concept of the data breach lifecycle. By incorporating this lifecycle and utilizing a novel self-exciting marked point process model, we enhance our understanding of the temporal dynamics of data breaches. Applications in insurance industry are also presented.

个人简介：

赵鹏，江苏师范大学二级教授、副校长。主要从事可靠性统计和网络可靠性等领域的研究工作。先后入选基金委国家优青项目、杰青项目。担任期刊Commun. Stat.副主编（AE）、《应用概率统计》《数理统计与管理》编委。曾获江苏省“双创人才”、江苏省数学成就奖、江苏省教学成果一等奖（第一完成人）等。担任中国现场统计研究会大数据统计分会副理事长、中国数学会理事、概率统计分会常务理事等。