Journal of Soft Computing and Artificial Intelligence

Öz

Highway traffic injuries are a major public health problem for all nations. As it is seen in the whole world, also in Türkiye, road traffic crashes are among the ones that cause death. As a result, road traffic congestion and fatalities represent a significant cost to national economies. The compulsory motor vehicle liability insurance is one of the most common types of insurance, both because it is compulsory and because the number of motor vehicles in Türkiye is very high. Therefore, estimation of the traffic insurance gross premiums (TIGP) problem is being an important problem for Türkiye as well as the other countries. In this study, in order to make some proper TIGP estimations for Türkiye, three different SMA methods such as SMA-Linear, SMA-Quadratic and SMA-Exponential are proposed. In the experiments, the population, number of vehicles and number of accidents and the observed TIGP historical data records of Türkiye taken from Turkish statistical institute (TUIK) and Turkish insurance association (TSB) for the years (2009-2020) have been used. First, the models are constructed using the SMA-Linear, SMA-Quadratic and SMA-Exponential methods, and then the methods based on the SMA-Linear, SMA-Quadratic and SMA-Exponential models are used to estimate the TIGP values for the years (2009-2020). According to the experimental results, SMA-Quadratic methods is produced better or comparable performance on the problem dealt with this study in terms of solution quality and robustness.