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：