报告题目:Modeling Malicious Hacking Data Breach Risks
Malicious hacking data breach has caused millions of dollars of financial losses each year, and more and more companies seek for the cyber insurance coverages. The lack of suitable statistical approaches for scoring the breach risks has become an obstacle in the insurance industry. We propose a novel frequency-severity model for analyzing the hacking breach risks from the individual company level which would be valuable for the underwriting purpose. We discover that the breach frequency can be modeled by a hurdle-Poisson model which is different from the negative binomial model used in the literature, and the breach severity shows a heavy tail which can be captured by a nonparametric-GPD model. We further discover a positive nonlinear dependence between the frequency and severity, which is also accommodated in our model. Both the in-sample and out-of-sample studies show that the proposed frequency-severity model by accommodating the nonlinear dependence has satisfactory performances, and is superior to the other models including the independence frequency-severity and Tweedie models.
徐茂超教授于2010年在美国波特兰州立大学获得统计学博士学位。现为美国伊利诺伊州立大学数学系副教授，担任美国CloudCover公司网络保险顾问。曾入选国家青年千人计划；获伊利诺伊州立大学杰出科研奖。研究方向主要包括应用统计、网络风险管理、网络安全保险等，近年来的相关研究成果发表于统计学、保险精算领域，如Technometrics，IEEE Transactions on Information Forensics and Security，IISE Transactions，Insurance: Mathematics and Economics等国际著名期刊上。
版权所有：南京航空航天大学 ALL RIGHTS RESERVED 苏ICP备05070685号