CAJ | 학술논문

论文将从多目标优化的角度出发，结合LDA的第一个目标（最大化类间方差）和SVM 的第二个目标（最小化经验风险），构造一个新的最大类间方差和最小经验风险（MVE）数据分类模型。由于该模型是一个非凸规划模型，论文使用凹凸规划（CCCP）来进行求解。为了验证论文提出的数据类模型，对人工和真实的数据挖掘实验室数据（UCI）数据集进行分类实验测试。实验结果表明该数据分类模型有效性的。
논문장종다목표우화적각도출발，결합LDA적제일개목표（최대화류간방차）화SVM 적제이개목표（최소화경험풍험），구조일개신적최대류간방차화최소경험풍험（MVE）수거분류모형。유우해모형시일개비철규화모형，논문사용요철규화（CCCP）래진행구해。위료험증논문제출적수거류모형，대인공화진실적수거알굴실험실수거（UCI）수거집진행분류실험측시。실험결과표명해수거분류모형유효성적。
For multi-objective optimization ,the LDA's first goals(maximum inter-class variance) and the SVM second goal(minimum the empirical risk) are combined for configuring a new MVE data classification model .It is a non-convex pro-gramming model and CCCP programming is utilized .In order to verify the data classification model presented in this paper , some data classification experiments are carried based on artificial and real UCI datasets .The experimental results show that the data classification model is effective .