支持向量机决策树在隐患预警模型中的应用  

Risk Early-Warning Model Based on SVM Decision Tree

作　　者：闫晓静[1,2] 于放[2] 孙咏[2] 肖卡飞 王嵩[2] YAN Xiao-Jing YU Fang SUN Yong XIAO Ka-Fei WANG Song(University of Chinese Academy of Sciences, Beijing 100049, China Shenyang Institute of Computing Technology, Chinese Academy of Sciences, Shenyang 110168, China)

机构地区：[1]中国科学院大学,北京100049 [2]中国科学院沈阳计算技术研究所,沈阳110168

出　　处：《计算机系统应用》2017年第2期212-216,共5页Computer Systems & Applications

摘　　要：危化企业的安全监控数据具有社会价值,对安全隐患进行实时精确的预测是预警研究的热点,本文从人、设备、环境和管理四个维度出发,对安全生产隐患预警的相关指标进行分析,构建隐患预警指标体系,在此基础上,构建了自底向上的基于支持向量机的决策树多分类预警模型,实现对安全等级的的准确分类并用于预警未来的安全生产状态,通过与自顶向下的多分类模型比较,证实本文所采用的预警模型具有较好的实时性和精确度,满足对预警模型的基本要求.The security monitoring data of Dangerous chemicals business has great social value, especially real-time accurate prediction of the security risk has become a hot warning research. From the view of four dimensions which are people, equipment, the environment and management, this article analyzes the relevant indicators of safety hazards warning, constructs the bottom-up decision tree based on multi-classification SVM warning model, constructs a bottom-up decision tree SVM multi-classification model based on early warning, to achieve the security level of accurate classification and for future production safety status warning. By comparison with more top-down classification model, it confirms that early warning model used in this paper has better performance in real-time and accuracy, and meets the basic requirements of early warning models.

关 键 词：预警模型 支持向量机 决策树 

分 类 号：F426.7[经济管理—产业经济] TP18[自动化与计算机技术—控制理论与控制工程]