[Response] Thanks for your comments. Because the peak height value at 3471cm-1 was obtained after subtracting the original spectrum of fresh oil, therefore the difference is small, but it still has the meaningful after the statistical analysis of one-way analysis of variance (ANOVA) and Duncan’s multiple range tests.
29. Did the author include the absorbance of only 3 peaks in the X matrix (i.e., predictors)? If that is the case, what is the justification for using ANN with limited number of predictors? Why didn’t the author choose other modeling algorithm such as multi-linear regression?