一种基于粗糙集和欧式距离的手机故障案例匹配算法

A Failure Case Matching Algorithm Based on Euclidean Distance and Rough Sets

本文研究在云环境中手机故障案例检索时的案例匹配问题,提出了一种基于粗糙集和欧式距离的案例相似度匹配算法CMARE(Case Matching Algorithm based on Rough sets and Euclidean distance)。首先云计算平台收集手机故障参数,根据参数构建粗糙集信息表,利用粗糙集求出信息表里各案例特征参数的属性客观权重值并结合专家经验给出综合的属性权重值,最后利用此权重值和欧式距离计算案例间的相似度,找出案例库中与新案例最相似的案例。该算法的创新之处在于确定案例属性权重值时基于数据本身和人工经验,避免了过分依靠人工经验知识设定属性权重的不足;与规则推理方式不同,本文使用案例推理的方式。仿真实例说明了算法的有效性。

Study the mobile phone fault case retrieval in a cloud environment when matching cases,we propose a similarity matching algorithm CMARE （Case Matching Algorithms based on Rough sets and Euclidean distance）.First,cloud computing platform collects mobile phone fault parameters,and then it builds a Rough sets information table according to these parameters.Second,property weight is computed based on objective weight of feature parameters of each case in the table calculated by Rough sets as well as in the light of expertise.Finally,with the weight and Euclidean distance,the degree of similarity between cases are figured out and then the most similar one from the case base is chased down.This algorithm is based on the date and experience,avoiding the disadvantage of being overly dependent on artificial expertise.This algorithm is based on the date and experience,avoiding the disadvantage of being overly dependent on artificial expertise.The simulation shows the effectiveness of the algorithm.