摘要
针对目前交通量预测中所广泛采用的基于经验风险最小化的BP网络易于陷入局部最优解等缺点,一种新的预测方法——基于结构风险最小化的SVMR交通量预测模型经实践证明能够较好地解决道路交通量预测问题。
BP network based on experimental risk minimization is easy plunging into local minimization in forecasting the city road traffic flow. Aiming at this defect, a new prediction method SVMR model based on structural risk minimization is proved being useful in short-time forecasting of the city road traffic flow.
出处
《交通标准化》
2006年第9期158-161,共4页
Communications Standardization
关键词
支持向量机
交通量预测
回归
support vector machine
traffic flow forecasting
regression
