Transport Problems (Jan 2010)
Neural classifier of the communication damage size being a result of collision of vehicles in road traffic
Abstract
In the article the results of the attempts of MLP neural network application to define the size of a communication damage being the result of a road collision were presented. The size of the damage was used as a research parameter defined by the coefficient dependent on the cost of repair of the damaged vehicle and its market value. The elements of the damage mechanism determining the way of damage qualification were the inner factors of the system, that is; the technical features of the vehicles, the character features of the drivers, the influence of the weather conditions and the location of the event in time and space. The research was conducted on one thousand cases reported for liquidation in Silesian branch of one of the insurance companies. In the conducted research the working of the neural networks with the limited input data was checked.