【摘要】 为了提高复杂产品高维不平衡质量特性数据集关键质量特征识别效率,提出CEM-IG识别方法.通过调整CEM(classification EM algorithm)算法的K值输出不同的聚类结果,消除冗余样本后作为IG(information gain)算法的输入,并以IG作为判别质量特性重要程度的标准构建识别模型,最终输出最优关键质量特性集.算例结果表明,该方法将CEM的缺失值处理能力和IG的不相关特性筛选能力优势互补,能够有效降低不平衡和高维度带来的负面影响,正确识别产品关键质量特性.更多还原

【Abstract】 In order to improve the efficiency of critical-to-characteristics identification in high-dimensional imbalance data sets for complex product,the CEM algorithm is integrated with IG algorithm,that is,adjusting the K-values in CEM algorithm to get the different clustering results which are inputted for after step,obtaining the recognition model based on CTQ features optional set formed by the IG as to the standard for the discriminate importance of quality characteristics,and selecting the optimal CTQ sets after tests.The results showed that,this method combined the missing data processing power of CEM and irrelevant characteristics screening power of IG successfully,and can reduce the negative effects effectively from imbalances,and can high-dimension and correctly identify the CTQ.更多还原