Comparative Study of Sugarcane Average Unit Yield Prediction With Genetic BP Neural Network Algorithm

摘要：

Because sugarcane average unit yield was affected by multipk factors in its growth and its inherent law was lack of external correlation data mining, the precise of the prediction method was low. Recently, the adaptive of modern intelligent genetic neural network algorithm for multi-factor effect has been strong, and the prediction accuracy has been high, but with wluch in sugarcane average unit yield prediction the researches are few. In this paper, based on the characteristics of external factor variable, the input multiple factors of sugar varieties, weather, etc. are optimized by multiple regression model, and the weight and threshold and the network structure of neural network modd are optimized by the genetic model of SGA /IGA, etc., which improves the adaptive fitness of genetic BP neural network model. Moreover, an example comparison of this algorithm and the gray linear system, S-BP, SGA-BP, IGA-BP on sugarcane average unit yield prediction is made. The results show that the integrated predichon accuracy and effectiveness of the improved genetic BP (IGA-BP) algorithm model on sugarcane average unit yield is optimaL This research provides a means of accurate prediction for sugarcane market price in significant fluctuations.

关键词： sugarcane genetic BP neural network algorithm average unit yield prediction multiple regression model

作者: Yong-Chun Xu Shi-Quan Shen Zhen Chen

作者单位: Department of Computer Science Guangdong Polytechnic Institute Guangzhou, P.R.China College of Engineering South China Agricultural University Guangzhou, P.R.China

会议类型: 国际会议

会议名称: The 2nd IEEE International Conference on Advanced Computer Control(第二届先进计算机控制国际会议 ICACC 2010)

会议地点: 沈阳

会议语种:英文

页码: 340-343

在线出版日期: 2010-03-27（万方平台首次上网日期，不代表论文的发表时间）