Abstract: Traffic flow prediction was analyzed from characteristics; the method of K-nearest neighbor was applied in the non-parametric regression for short-term forecast of road traffic flow; the road traffic-related data of Guiyang City was employed to validate the model． It proves that the K-nearest neighbor nonparametric regression model can predict the shortterm traffic flow more accurately and it also reveals that the feasibility of the methods is used in short-term traffic flow prediction．

Key words: short-term traffic flow prediction, non-parametric regression, K-nearest neighborhood

摘要： 从分析短时交通流特性入手，利用非参数回归中K近邻的方法，对道路交通流量进行短期预测;采用贵阳市道 路交通流量的实际数据进行验证。结果表明:K近邻非参数回归预测模型能较为准确的进行道路短期交通流预测， 该方法可用于短期交通流预测。

关键词: 短期交通流预测, 非参数回归, K近邻