摘要
在模糊优选BP神经网络模型上,加入反馈信号及偏差单元,生成递归神经网络.其基本思想是通过对前馈神经网络中加入附加的、内部的反馈通道来增加网络本身处理动态信息的能力.将该模型用于水电投资项目的评价,所得结果与其它方法所得结果相比较,对该方法用于水利水电投资项目的评价的有效性进行了说明.
Recurrent neural network is generated by adding the feedback and the bias element to BP neural network model of fuzzy optimization, whose basic idea is to increase the network＇s ability to process dynamic information through adding extra and internal feedback channels to feedforward neural network. This paper compares the results from the application of the model in hydropower investment projects with those from other method, and explains the effectiveness of using this model in hydropower investment projects.
出处
《重庆工学院学报》
2006年第2期138-142,共5页
Journal of Chongqing Institute of Technology
