装备质量数据离散化方法  

Discretization method for equipment quality data

作　　者：李馥林 孟晨 王成 范书义 LI Fulin;MENG Chen;WANG Cheng;FAN Shuyi(Department of Missile Engineering,Shijiazhuang Campus of Army Engineering University,Shijiazhuang 050003,China)

机构地区：[1]陆军工程大学石家庄校区导弹工程系,石家庄050003

出　　处：《兵器装备工程学报》2023年第9期143-148,共6页Journal of Ordnance Equipment Engineering

基　　金：国家自然科学基金项目(61501493)。

摘　　要：数据挖掘技术已经成为一种利用装备数据资源获取知识的重要手段,数据预处理是装备质量信息分析的重要环节。为解决数据类型不适应数据挖掘方法的问题,提出了一种装备质量数据离散化方法。对经典类别属性最大相互依赖算法的原理和流程进行了介绍,分析了存在的问题,提出了改进方法;引入了粗糙集理论和属性分辨率,限制了过度离散化;提出了属性重要性评价方法,减少了信息损失。对比实验结果表明,所提出的方法具有优越性,能提高数据离散化效果。以某型装备为例,将所提方法应用于关联规则挖掘之前的数据预处理,获得了与装备寿命周期内质量变化规律相关的知识,验证了该方法的有效性。Data mining technology has become an important means of using equipment data resources to obtain knowledge.Data preprocessing is an important part of equipment quality information analysis.In order to solve the problem that data types do not adapt to data mining methods,a discrete method for equipment quality data is proposed.The principle and process of the classical class-attribute interdependency maximization algorithm are introduced,the existing problems are analyzed,and the modification method is proposed.Rough set theory and attribute resolution are introduced to limit over discretization,and attribute importance evaluation method is proposed to reduce information loss.The comparative experimental results show that the method proposed in this paper has advantages and can improve the effect of data discretization.Taking a certain type of equipment as an example,the method proposed in this paper is applied to the data preprocessing before association rule mining,and the knowledge related to the rule of quality change in the equipment life cycle is obtained,which verifies the effectiveness of the method.

关 键 词：装备质量 数据离散化 数据挖掘 数据预处理 关联规则 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]