26 February 2013 8 6K Report

I am using vis-NIR to measure soil samples and calibrate spectral data to predict unknown samples (those of samples have reference value from lab chemical analysis). After I get prediction results from NIR measurement, I want to use prediction error (such as RMSEP or SEP) from NIR to compare error from lab chemical analysis. My question is does it have any principal for this kind of comparison, and how to calculate lab error?

More Yi Peng's questions See All
Similar questions and discussions