|
| 应用数据挖掘算法预测台风条件下PM2.5质量浓度 |
| Prediction of PM2.5 during the Typhoon Using Data Mining Algorithms |
| |
| DOI： |
| 中文关键词: PM2.5 台风 数据挖掘算法 预测 环境空气 |
| 英文关键词: PM2.5 Typhoon Data mining algorithms Prediction Ambient air |
| 基金项目: |
|
| 摘要点击次数: 854 |
| 全文下载次数: 0 |
| 中文摘要: |
| 利用6种数据挖掘算法对2012年台风和污染物浓度等数据建模，建立台风条件下的气象因素和环境空气中PM2.5的关系模型。通过对2013年数据的测试，表明6种数据挖掘方法模型的预测值与实测值具有很好的一致性。6种算法均能实现实时预测，其中Bagging算法预测结果的相关性整体最好，而对于测试样本，M5Rules算法最好，SMOreg算法次之。以2013年的超强台风“海燕”的预测为例，讨论台风条件下PM2.5质量浓度的变化过程及对预测结果的解释。 |
| 英文摘要: |
| Based on six kinds of data mining algorithms, the relation models were established by the meteorological factors and PM2.5 concentrations in ambient air under the condition of typhoon in 2012. The models were applied to analyze the data in 2013, the results showed that the predicting values and measured values were generally consistent. All of the algorithms can be implemented in real time. Among all the algorithms, the Bagging algorithm was better than others. But for the test data in 2013, the M5Rules algorithm was the best and the SMOreg algorithm the second. The influence of typhoon on the concentration of PM2.5 and the explanation of predicting result were discussed. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
