【作者】 黄庚保； 张根保； 曾海峰； 王国强； 刘立堃；

【Author】 Huang Gengbao 1、2,Zhang Genbao 1,Zeng Haifeng 1,Wang Guoqiang 1,Liu Likun 1 1.Chongqing University mechanical engineering institute,Chongqing 400044,P.R.China 2.Guangxi university of technology,Guangxi,Liuzhou 545006,P.R.China

【机构】 重庆大学机械工程学院； 广西工学院；

【摘要】 由于有隐秘性质量参数的存在,致使质量水平测度过程中人为的出现输出结果总是处于良好,使计算机的学习能力不能显示优势。本文根据质量要素的客观性和评价群体的不同,提出了具有向量意义的三维质量测度指标表示方法,针对隐秘性参数所在的空间不同,设置不同的学习方法和计算途径,构成由复合线性矩阵与支持向量机组成的并联学习机方案,间接隔离出隐秘性参数的保护性信道。更多还原

【Abstract】 Because of the existence of hidden parameters for the quality,the output is always artificially in good result during the quality measurement process,and the computer can not show the advantages of learning ability.According to the objectivity of quality factors and different evaluation groups,a method of three-dimensional quality measurement indicator having vector significance is presented.In view of the different space of hidden parameters,different learning methods and ways of calculation are set,and the Parallel Learning Machine scheme consisting of a compound linear matrix and Support Vector Machine is formed,which indirectly isolates protective channel of hidden parameters.更多还原