节点文献
自组织特征映射神经网络在厄尔尼诺事件检验中的应用
The Application of the Self-Organizing Feature Map Neural Network in Testing the El Nino Event
【摘要】 对厄尔尼诺事件多因素成因进行了分析。利用自组织特征映射(SOFM)神经网络方法对1973~1994年的全球7级以上地震次数、日食条件、海温距平数据建立了SOFM网络检验模型。对1995~2000年厄尔尼诺事件进行了检验,检验的准确率为83.3%。
【Abstract】 The causes of the El Nino events was analyzed.The self-organizing feature map(SOFM) neural network forecasting model was built up according to the numbers of the earthquakes,the conditions of solar-eclipses and average of sea-temperature in 1973~1994 with the method of the SOFM network and examines the El Nino events happened in 1995~2000.Its accuracy rate is 83.3%.
【关键词】 自组织特征映射;
人工神经网络;
厄尔尼诺;
日食;
海温;
地震;
【Key words】 self-organizing feature map; artificial neural network; El Nino; solar-eclipse; sea-temperature; earthquake;
【Key words】 self-organizing feature map; artificial neural network; El Nino; solar-eclipse; sea-temperature; earthquake;
【基金】 国家“863”计划项目(2001AA135120-2)
- 【文献出处】 吉林大学学报(地球科学版) ,Journal of Jilin University(Earth Science Edition) , 编辑部邮箱 ,2006年04期
- 【分类号】P732
- 【被引频次】2
- 【下载频次】82