摘要
采用集成学习的思想,提出了一种基于集成特征选择的森林火灾风险评估方法。以特征选择方法的多样性和独立性为考量,选择了15种特征选择器并利用差异度进行筛选,获得异质选择器集合,进而得到特征子集集合。其次,利用各特征子集分别构建基于BP神经网络的森林火灾风险评估模型,并依据模型准确度筛选林火重要影响因子,构建最优森林火灾风险评估模型。结果表明,该算法准确度为85.96%,具有良好的泛化能力,可实现对森林火灾风险的有效评估。
This paper proposes a forest fire risk assessment method based on ensemble feature selection.Considering the di-versity and independence of algorithms,15 kinds of feature se-lect algorithms are selected to form the heterogeneous selectors based on their difference.By using the feature select algo-rithms,a feature subset set is obtained.And then a forest fire risk assessment model is constructed based on BP neural net-work by using each feature subset.The important factors of for-est fires are selected based on the accuracy of neural network to construct the optimal forest fire risk assessment model.The re-sults show that the accuracy of the algorithm proposed is 85.96%.The proposed model has good generalization ability and can as-sessforestfireriskeffectively.
作者
周文涛
张皓
陈维捷
周游
Zhou Wentao;Zhang Hao;Chen Weijie;Zhou You(School of Electrical and Information Engineering,Changsha University of Science and Technology,Hu'nan Chang-sha 410114,China;State Grid Shandong Electric Power Com-pany Electric Power Research Institute,Shandong Ji'nan 250003,China)
出处
《消防科学与技术》
CAS
北大核心
2022年第12期1727-1731,共5页
Fire Science and Technology
基金
长沙理工大学学术学位研究生科研创新项目(CX2020SS56)
长沙理工大学研究生实践创新与创业能力提升项目(SJCX202045)
关键词
集成特征选择
森林火灾
BP神经网络
风险评估
ensemble feature selection
forest fire
BP neural network
risk assessment
作者简介
周文涛(1997-),男,长沙理工大学电气与信息工程学院硕士研究生,主要从事森林火灾风险分布评估方面的研究,湖南省长沙市(天心区)万家丽南路2段960号工科一号楼,410114。
