Development of Optimum Neural Network Model for Rainfall Prediction System

摘要：

Neural Network models are increasing interest in various knowledge management researches and application domains such as weather prediction and data mining from heterogeneous data sets. However, it is difficult to determine the best neural network architecture for prediction domains since it is not immediately obvious how many input or hidden nodes are used in the model. In this paper, the better performance neural network model is proposed and the system has tested 100 cases by changing the number of input and hidden nodes from 1 to 10 nodes, respectively and only 1 output node is used in all cases. Results show that 3-inputs-10 hiddens-1 output neural network model gives the better prediction result for monthly rainfall in Myanmar.

作者: Khaing Win Mar Thinn Thu Naing

作者单位: University of Computer Studies, Yangon, Myanmar

会议类型: 国际会议

会议名称: The 9th International Symposium on Knowledge and Systems Sciences,The 4th Asia-Pacific International Conference on Knowledge Management(第九届国际知识与系统科学学术年会暨第四届亚太国际知识管理年会)

会议地点: 广州

会议语种:英文

页码: 405-409

在线出版日期: 2008-12-11（万方平台首次上网日期，不代表论文的发表时间）