摘要
本文成功地将神经网络应用于强对流天气的短时预报。此外,我们使用APCA算法对天气样本的参数进行了优选,压缩了参数个数,在预报成功率基本不变的情况下,减轻了数据的采集和整理的工作量。
In this paper, BP Network is applied in the prediction of intense convective weather. The correct prediction rate obtained is higher than the result obtained with stochastic analysis technique. In addition, Asymmetric Principal Component Analysis is used to select the parameters of the weather patterns. With the almost same correct rate, the number of the parameters is decreased. This work decreases the difficulty of data collection and preprocessing.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1998年第3期323-327,共5页
Pattern Recognition and Artificial Intelligence
关键词
天气预报
BP网络
特征选择
异主元分析
Weather Prediction, BP Network, Feature Selection, Asymmetric Principal Component Analysis
