论文部分内容阅读
建立了调用NEWRB函数的正规化网络RN和基于K-means聚类的广义网络GN的两种RBF‘神经网络的工程造价预测模型,以55个厦门市工程造价案例进行实证分析.结果表明:当调用NEWRB函数构建RBF模型时,其性能主要取决于分布宽度,而基于K-means聚类的RBF神经网络主要取决于重叠系数和隐含层节点数;基于广义网络GN的RBF神经网络模型的训练效果较差,但学习速度更快、预测精度更高.
The construction cost forecasting model of two kinds of RBF neural networks, which are the normalized network RN which calls NEWRB function and the generalized network GN which is based on K-means clustering, is established, and an empirical analysis is carried out with 55 cases of construction cost in Xiamen City.The results show that when When RBF model is constructed by calling NEWRB function, its performance depends on the distribution width, while the RBF neural network based on K-means clustering mainly depends on the overlap coefficient and the number of nodes in hidden layer. RBF neural network model based on GNN training Less effective, but learning faster, higher prediction accuracy.