Outliers and Preprocessing

Log Out | Topics | Search Moderators | Register | Edit Profile

Author Message   Hongtu Xie (tonali) Junior Member Username: tonali Post Number: 7 Registered: 2-2010 Posted on Monday, March 08, 2010 - 8:14 am:    ha, I'm reading chemometrics and NIR books. Now I'm still confused, maybe in the near future, I can make sense.   Hongtu Xie (tonali) New member Username: tonali Post Number: 4 Registered: 2-2010 Posted on Thursday, February 25, 2010 - 10:32 am:    Hello, all I was confused by the outlier detection for a longtime. When I use software (e.g.OPUS, Unscrampbler etc.) to calibrate soil sample and made a model. How to evaluate the outliers(e.g. If I removed some outliers and run the calibration again, and other outliers appeared)? Mahalanobis distance? PCA? F values? etc. Are there any standards for these? The preprocessing confused me too. So many preprocessing method, how to select and what's the sequence? I read some posts about preprocessing here, but didn't make sense. I don't know which preprocessing method should be used. And some authors didn't remove outliers and didn't preprocess spectra before analysis. Are these depend on experience? Could you give me some suggestions, comments or maybe recommend some references or books. Many thanks!