In order to select the efficient input variables of adaptive ncuro-fuzzy infence system （ANFIS）during the prediction anthropometric dimenions, grey incidence （GI） analysis, as a mastic method that ranks the sequen...In order to select the efficient input variables of adaptive ncuro-fuzzy infence system （ANFIS）during the prediction anthropometric dimenions, grey incidence （GI） analysis, as a mastic method that ranks the sequence of of lots of variables in complicated factors has been applled.According to the prediction accuracy （A） between the predicted values and actual measured values, the ANFISG1 model with the parameters selected by using the GI analysis were more correct and effective than those done by multiple regression model and the model with input parmeters nonelected. The model prediction accuracy △Regrauskn= 0.804 7〈 △ANE3CI=0.9725, which proves the nodel with few parameters is more correct and effective than the other merits.展开更多
基金Shanghai Board of Education Scientific Research Projects (No.106N2013)
文摘In order to select the efficient input variables of adaptive ncuro-fuzzy infence system （ANFIS）during the prediction anthropometric dimenions, grey incidence （GI） analysis, as a mastic method that ranks the sequence of of lots of variables in complicated factors has been applled.According to the prediction accuracy （A） between the predicted values and actual measured values, the ANFISG1 model with the parameters selected by using the GI analysis were more correct and effective than those done by multiple regression model and the model with input parmeters nonelected. The model prediction accuracy △Regrauskn= 0.804 7〈 △ANE3CI=0.9725, which proves the nodel with few parameters is more correct and effective than the other merits.
