基于广义回归神经网络的交通事故预测

Traffic accident prediction based on GRNN

在进行交通事故统计的基础上，运用广义回归神经网络理论，建立交通事故预测模型，对2000-2003年的交通事故死亡人数和经济损失进行外推预测．计算结果表明，该模型预测和外推精度高，可用于交通事故预测．

Based on the analysis of road trattic acctdent, a set ot general regression network model was developed for forecasting traffic accident. The GRNN model was applied to extrapolate the number of death and economy lose of 2000-2003. The result shows that the model is precise and feasible.