Γραμμικά μοντέλα παλινδρόμησης για δεδομένα ασφαλιστικών αποζημιώσεων
Linear regression models for insurance data
KeywordsΓραμμική παλινδρόμηση ; Λογιστική παλινδρόμηση ; Γραμμικά μοντέλα ; Ασφαλιστικές αποζημιώσεις ; Αντασφάλιση
In many statistical applications, we encounter the problem of studying the relationship of two or more random variables. This relationship can be studied with appropriate regression models, which are widely used today in business administration, economy, health, insurance, biology and social sciences. In statistics, regression analysis is a process for quantifying and studying the relationship between different variables. It contains many techniques for the modeling and analysis of these variables, while it usually focuses on the linear relationship between a dependent variable and one or more independent variables. Then, besides the description of the above method, this diploma thesis presents definitions and techniques of risk theory for the study of insurance indemnities, as well as analyses the cut-off point of reinsurance that an insurance company could choose, in order to «protect its funds». Continuous and accurate risk assessment in the insurance industry to ensure that there are sufficient reserves, is rooted in the 1970s and redefined in the mid-1990s.