Although the increasing expansion of traffic in cities has increased economic and welfare benefits, it has, on the contrary, increased the number and severity of traffic accidents. Reducing the number of victims and injuries caused by road accidents in any common moral-value system is urgent and inevitable. In this way finding effective factors on the severity of road injuries can be considered as an effective step towards achieving the values. Finding effective factors on severity of injuries, with emphasis on statistical efficacy of effective policy-making factors, will be used as an appropriate tool in the middle level of road safety management. Accident prediction results using MATLAB software in selected roads showed that although this model, by choosing the appropriate calibration factor and using the appropriate parameters and high precision, can produce good outputs, but the results are less accurate than the MLP. The statistical analysis of the observed values ​​and the predicted crash values ​​showed that their differences were not statistically significant at the 5% confidence level, and their results could be used to predict crashes and determine future conditions.