摘 要:种灾害的发生通常会伴随多种灾害同时出现,并对环境产生影响,对多灾种之间的内在关系进行分析,确定其领域知识结构,并基于知识领域对已有的单灾种灾害链预警模型进行分析,最终获得通用性较强且可重复使用的多灾种灾害链预警模型。这种方法不需要大量数据集,可以同时处理多个任务,实验表明,通过对已有单灾种预警模型的重用,预警的准确性得到一定提升。
关键词:深度学习;多灾种;预警模型;重用
DOI:10.19850/j.cnki.2096-4706.2022.09.043
基金项目:湖南省应急管理厅 2020 年度科技项目(2020YJ008)
中图分类号:TP18 文献标识码:A 文章编号:2096-4706(2022)09-0173-04
Design of Reusing Framework for Multi Disaster Chain Early Warning Model Based on Machine Learning
XIA Xu
(Hunan Vocational Institute of Safety Technology, Changsha 410151, China)
Abstract: The occurrence of one disaster will lead to multiple disasters, which will has an impact on the environment. The internal relationship among multiple disasters is analyzed, the domain knowledge structure is determined, based on the knowledge domain, the existing single disaster chain early warning model is analyzed, and finally a multi disaster chain early warning model with strong universality and reusability is obtained. This method does not need a large number of data sets, and can handle multiple tasks at the same time. Experiment results show that the accuracy of early warning can be improved by reusing the existing single disaster early warning models.
Keywords: deep learning; multi disaster; early warning model; reuse
参考文献:
[1] 张永利,张建平,任爱珠,等 . 多智能体的多灾种耦合预测模型 [J]. 清华大学学报(自然科学版),2011,51(2):198-203.
[2] 汪嘉俊,翁文国 . 多灾种概念辨析及灾害事故关系研究综述 [J]. 中国安全生产科学技术,2019,15(11):57-64.
[3] 郭增建,秦保燕 . 灾害物理学简论 [J]. 灾害学,1987(2):25-33.
[4] AGHAKOUCHAK A,HUNING L S,CHIANG F,etal.Howdonaturalhazardscascadetocausedisasters ? [J].Nature561,2018:458-460.
[5] 哈斯,张继权,郭恩亮,等 . 基于贝叶斯网络的草原干旱雪灾灾害链推理模型研究 [J]. 自然灾害学报,2016,25(4):20-29.
[6] ZHOU Z H.Learnware:onthefutureofmachinelearning [J].Frontiers of Computer Science,2016,10(4):589-590.
[7] DAI W Z,ZHOU Z H.CombiningLogicAlabductiona ndStatisticalInduction:DiscoveringWrittenPrimitivesWithHuma nKnowledge [EB/OL].[2022-01-02].https://www.docin.com/p-1957262116.html.
[8] ZHAO P,JIANG Y,ZHOU Z H.Multi-ViewMatrixC ompletionforClusteringwithSideInformation [J/OL].Advances in Knowledge Discovery and Data Mining,2017:403-415 [2022-01-02].https://link.springer.com/chapter/10.1007/978-3-319-57529-2_32#citeas.
作者简介:夏旭(1980—),女,汉族,湖南益阳人,教授,硕士,主要研究方向:机器学习、安全监控技术。