| 基于改进型BP算法的外债风险指标预测 |
| |
| 引用本文： | 陈雄华,林成德.基于改进型BP算法的外债风险指标预测[J].厦门大学学报(自然科学版),2001,40(5):1017-1021. |
| |
| 作者姓名： | 陈雄华 林成德 |
| |
| 作者单位： | 厦门大学自动化系, |
| |
| 摘 要： | 利用人工神经网络进行时间序列预测是一种较新的方法，它具有不需建立复杂的数学模型以及非线性映射能力强等优点。采用动量法和学习率自适应调整的改进型BP算法对外债风险的各项指标进行了非线性时间序列的预测。仿真结果表明神经网络模型对外债风险的各项指标预测的结果是准确可靠的。
|
| 关 键 词： | 外债风险 非线性时间序列预测 人工神经网络 BP算法 指标预测 动量法 学习率自适应调速 |
| 文章编号： | 0438-0479(2001)05-1017-05 |
| 修稿时间： | 2000年6月9日 |
Prediction of Foreign Debts Risk Indicators Based on Improved BP Algorithm |
| |
| CHEN Xiong hua,LIN Cheng de.Prediction of Foreign Debts Risk Indicators Based on Improved BP Algorithm[J].Journal of Xiamen University(Natural Science),2001,40(5):1017-1021. |
| |
| Authors: | CHEN Xiong hua LIN Cheng de |
| |
| Abstract: | Time series prediction, based on artificial neural network, is newly developed. It can approximate properly nonlinear mappings without building very complicated mathematical models. By using an improved BP algorithm, the risk indicators on foreign debts with nonlinear time series are predicted. The result s of the prediction turn out that the artificial neural network is very applicab le and rather exact in this domain. |
| |
| Keywords: | foreign debts risk artificial neural network nonlinear time series prediction BP algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
