| 姚作芳,刘兴土,杨飞,闫敏华.几种方法在粮食总产量预测中的对比[J].干旱地区农业研究,2010,28(4):264~268 |
| 几种方法在粮食总产量预测中的对比 |
| Comparison of several methods in grain production prediction |
| |
| DOI：10.7606/j.issn.1000-7601.2010.04.49 |
| 中文关键词: 主成分分析法 BP神经网络法 灰色预测法 逐步回归法 粮食产量 预测 |
| 英文关键词:principal component analysis neural network method GM(1,1)grey prediction model regression analysis model grain production prediction |
| 基金项目:中国科学院知识创新工程重大项目(KSCX1-YW-09-13)；农业气候资源评价与高效利用技术研究专项(GYHY200706030) |
|
| 摘要点击次数: 207 |
| 全文下载次数: 137 |
| 中文摘要: |
| 根据吉林省19782007年的农业数据,选定了14个影响农业生产的因素作为研究对象,分别采用了主成分分析法、BP神经网络法、灰色预测法和逐步回归分析法4种分析预测方法,通过SPSS和Matlab工具将原始数据进行处理,得到4种不同的预测模型,进而基于这4种模型对吉林省的粮食产量进行预测,并将各种预测产量和实际产量进行拟合分析。研究结果表明,拟合性最好的是BP神经网络法,其拟合确定性系数为0.899;其次是主成分分析法(拟合确定性系数为0.834)和逐步回归法(拟合确定性系数为0.787);拟合效果最差的是灰色预测法(确定性系数为0.744)。粮食总产量估算精度最高的是BP神经网络法,达到93.67%;其次是主成分分析法,为90.45%。 |
| 英文摘要: |
| In this study, four predicting methods, namely principal component analysis, BP neural network
method, GM(1,1) grey prediction model and regression analysis model, were used and compared for grain production prediction. Based on agricultural data from 1978 ~ 2007 of Jilin Province，fourteen affecting factors were chosen for anal-
ysis by SPSS and Matlab tools to build four different prediction models. And these models were also used to predict the
grain production in Jilin Province，and comparison was made of the R2 of each regression line and the prediction accuracy
of different models. The studied results showed that BP neural network method performed the best with regression R2 and estimation accuracy of 0.899 and 93.67%，principal component analysis method performed moderate grain production estimation results, with R2 of 0.834 and the accuracy of 90.45%，the next is stepwise regression analysis, while grey prediction method was the poorest. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| | |