基于决策树的水稻病虫害发生程度预测模型——以芜湖市为例
A prediction model for the occurrence degree of rice diseases and pests based on decision tree algorithm:A case study in Wuhu
付伟 1祝玉青 1司红君 1邹莹瑾2
作者信息
- 1. 芜湖市气象局,安徽 芜湖 241000
- 2. 无为市气象局,安徽 无为 238300
- 折叠
摘要
利用芜湖市1988-2022年7种水稻病虫害发生及种植面积数据,国家气候中心逐月大气环流和海温指数,基于C5.0决策树算法,构建一种以大气环流和海温指数作为预报因子的水稻病虫害发生程度的长期预测模型.结果表明,该模型可较好地预测未来一年芜湖市各种水稻病虫害发生程度,2022年7种病虫害发生程度预测准确率平均为85.7%,为水稻病虫害发生程度的预测提供一种有效的实用方法.
Abstract
Based on the C5.0 decision tree algorithm,a long-term prediction model for the occurrence of rice diseases and pests using atmospheric circulation and sea surface temperature(SST)index as predictive factors was constructed using the occurrence of seven types of rice pests and diseases and planting area data of Wuhu City from 1988 to 2022,and the monthly atmospheric circulation and SST index of National Climate Centre(NCC).The results show that these models can effectively predict the occurrence degree of various rice diseases and pests in Wuhu City in the next year.The average predicting accuracy of the occurrence degree of seven diseases and pests in 2022 is 85.7%,which provides an effective and practical method for predicting the occurrence degree of rice diseases and pests.
关键词
C5.0决策树算法/水稻病虫害发生程度/预测模型/大气环流和海温指数Key words
C5.0 decision tree algorithm/occurrence degree of rice pests and diseases/prediction model/atmospheric circulation and sea surface temperature index引用本文复制引用
基金项目
国家重点研发计划(2018YFD0300905)
科技助力经济2020重点专项气象行业项目(KJZLJJ202002)
出版年
2024
国家科技期刊平台
NSTL
万方数据