Abstract
Road accidents are a major concern to the society and for the individualsinvolved. All the consequences of road accidents motivate research to understand the factors behind the causalities and the effort to minimise the number of accidents. The causes of the road accidents were examined using two databases recording data about the road accidents and the road conditions at the scene of accidents in Southern Finland. Data are analysed in anexploratory fashion using an artificial neural network called the self-organising map. In addition, fatal accidents are classified using a Naive Bayes classifier. Fatality and injury risks are calculated for various accident conditions. For the covering abstract see ITRD E139491.