基于決策樹算法的江蘇省不同區(qū)域 短時強降水預報研究

doi:10.12306/2022jms.0052

首頁 > 過刊瀏覽>2022年第42卷第5期 >631-637. DOI:10.12306/2022jms.0052

基于決策樹算法的江蘇省不同區(qū)域 短時強降水預報研究
DOI:
                        10.12306/2022jms.0052
                    
CSTR:
                        
                    
作者:
                        
                        
                    
作者單位:
作者簡介:
通訊作者:
中圖分類號:P409
基金項目:江蘇省科協(xié)青年科技人才托舉工程項目(TJ-2021-21);連云港市重點研發(fā)計劃(SF2132);江蘇省氣象局青年基金項目(KQ202123);連云港市"521工程"科研項目(LYG06521202138);江蘇省氣象局科研項目(KZ202107)

Study on short-term heavy precipitation forecast in different regions of Jiangsu basedon decision tree algorithm

Author:

Affiliation:

Fund Project:

摘要

圖/表

訪問統(tǒng)計

參考文獻

相似文獻

引證文獻

資源附件

文章評論

摘要:

采用2000—2019年13個地級市氣象站地面觀測站點觀測資料以及ERA5再分析資料,基于機器學習中的經(jīng)典的C4.5算法對江蘇省不同區(qū)域是否出現(xiàn)短時強降水建立氣象要素預報模型。結(jié)果表明:基于C4.5算法的決策樹預測模型能夠較為直觀準確的對江蘇省不同區(qū)域是否發(fā)生短時強降水進行預測,并且該決策樹模型具有較高的泛化能力。決策樹模型利用各區(qū)域總樣本的前15 a數(shù)據(jù)樣本進行自學習,學習準確率在淮北地區(qū)為89.70%,在江淮之間地區(qū)為87.89%,在長江以南地區(qū)為87.88%,利用各區(qū)域剩余5 a樣本對該決策樹模型的泛化能力進行測試,測試準確率在淮北地區(qū)為85.73%,在江淮之間地區(qū)為83.39%,在長江以南地區(qū)為93.92%。

Abstract:

Based on the observation data from 13 prefecture-level meteorological stations from 2000 to 2019 and associated ERA5 reanalysis data, Based on the classic C4.5 algorithm in machine learning, established a meteorological element prediction model for short-term heavy precipitation forecast in different regions of Jiangsu. The results show that the decision tree prediction model based on C4.5 algorithm can intuitively and accurately predict the occurrence of short-term heavy precipitation in different regions of Jiangsu, and the decision tree model has a high generalization ability. The self-learning process is conducted based on the first 15 years of the total samples of each region. The learning accuracy is 89.70% in the Huaibei region, 87.89% in the region between the Yangtze River and Huaihe River, and 87.88% in the region south of the Yangtze River, respectively. The generalization ability of the decision tree model is tested by using the remaining 5 years samples in each region, and the test accuracy is 85.73% in the Huaibei region, 83.39% between the Yangtze River and Huaihe River, and 93.92% in the region south of the Yangtze River, respectively.

參考文獻

相似文獻

引證文獻

引用本文

史達偉,沈陽,馬晨晨,董京銘,顏佳任.基于決策樹算法的江蘇省不同區(qū)域短時強降水預報研究.氣象科學,2022,42(5):631-637 SHI Dawei, SHEN Yang, MA Chenchen, DONG Jingming, YAN Jiaren. Study on short-term heavy precipitation forecast in different regions of Jiangsu basedon decision tree algorithm. Journal of the Meteorological Sciences,2022,42(5):631-637

復制

文章指標

點擊次數(shù):
下載次數(shù):
HTML閱讀次數(shù):
引用次數(shù):

歷史

收稿日期:2022-02-07
最后修改日期:2022-05-03
錄用日期:
在線發(fā)布日期: 2022-11-03
出版日期:

文章二維碼

抽查舔水白紧大视频,冲喜娘子超甜哒激情五月,亚洲黄色图片偷拍网站,亚洲三级黄片免费播放,91在线观看喷潮红桃,日韩精品在线观看一区二区,久久久久久久久久久久黄色片子,日产精品99久久久久久久久,欧洲亚洲另类图片小说

引用本文

分享

相關視頻

文章指標

歷史

文章二維碼

抽查舔水白紧大视频,冲喜娘子超甜哒 激情五月,亚洲黄色图片偷拍网站,亚洲三级黄片免费播放,91在线观看喷潮红桃,日韩精品在线观看一区二区,久久久久久久久久久久黄色片子,日产精品99久久久久久久久,欧洲亚洲另类图片小说

引用本文

分享

相關視頻

文章指標

歷史

文章二維碼

抽查舔水白紧大视频,冲喜娘子超甜哒激情五月,亚洲黄色图片偷拍网站,亚洲三级黄片免费播放,91在线观看喷潮红桃,日韩精品在线观看一区二区,久久久久久久久久久久黄色片子,日产精品99久久久久久久久,欧洲亚洲另类图片小说