Markov-Based Cyber Insurance Model for Dynamic Communication Network
Yeftanus Antonio (a), Sapto Wahyu Indratno (a,b*), and Suhadi Wido Saputro (c)

a)Statistics Research Division, Institut Teknologi Bandung, Bandung, West Java, Indonesia
b) University Center of Excellence on Artificial Intelligence for Vision, Natural Language Processing & Big Data Analytics (U-CoE AI-VLB), Institut Teknologi Bandung, Bandung, West Java, Indonesia
*sapto[at]math.itb.ac.id
c) Combinatorial Mathematics Research Division, Institut Teknologi Bandung, Bandung, West Java, Indonesia

Abstract

We present a Markov-based cyber insurance model for dynamic networks in this paper. The topology of the communication network has a critical role in the pricing or ratemaking of cyber insurance products. Static networks have been introduced with pre-existing models, obviating the need for the model to account for changes in structure over time. A dynamic network is an actively evolving network. Thus, the adjacency matrix is a time-dependent representation of the network or graph. To strengthen our findings, we simulate cyber incidents using the characteristics of a real temporal email network. These properties were adopted in a small hypothetical network or a random network for this study. Risk theories on dynamic networks using the assumption of random network connections are also obtained. The numerical studies obtained have also confirmed these results. This technique can be used to explain variations in network structure and their impact on premiums. The premium generated by this technique is sufficiently high to allow for the effects of structural variability across time.

Keywords: Cyber Insurance, Premium, Dynamic Network, Markov-Based Model

Topic: Mathematics