Management of cyber risks has become an important issue for any large company. However, academic research on quantitative stochastic modelling of cyber risk is virtually absent and there are no commonly accepted models and methods. This project focuses on the development of actuarial models for the quantification of cyber risk losses and the identification of key risk factors. Specifically, the project will quantify appropriate models for the frequency and severity of cyber risk losses and their dependence on risk indicators such as quality of software systems, transaction volume, country, company size available in the analysed dataset. To accomplish this, the team will develop nonlinear regression frameworks for frequency and severity parameters using machine learning and extreme value theory methods.
|Effective start/end date||1/06/20 → 1/06/21|