Validation NIR Model Log Out | Topics | Search
Moderators | Register | Edit Profile

NIR Discussion Forum » Bruce Campbell's List » Chemometrics » Validation NIR Model « Previous Next »

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

Pierre Dardenne (dardenne)
Senior Member
Username: dardenne

Post Number: 53
Registered: 3-2002
Posted on Thursday, December 02, 2010 - 3:04 pm:   

Hi,

concerning NIR validation, there is interesting information in ISO/12099.

Pierre
Top of pagePrevious messageNext messageBottom of page Link to this message

Howard Mark (hlmark)
Senior Member
Username: hlmark

Post Number: 366
Registered: 9-2001
Posted on Thursday, December 02, 2010 - 9:19 am:   

Fernando - there is another, fairly new, ASTM standard, E2617 "Standard Practice for Validation of Empirically Derived Multivariate Calibrations" that should address some of your questions. It gives more information, and also specifies more stringent conditions than E1655 does, to ensure proper validation.

Since it's new, it may still be undergoing revisions, so if you still have questions after reading E2617, send another question to this discussion thread and the primary authors of the Standards can address them (and possibly revise the standard based on your comments).

Howard

\o/
/_\
Top of pagePrevious messageNext messageBottom of page Link to this message

Fernando Morgado (fmorgado)
Junior Member
Username: fmorgado

Post Number: 6
Registered: 12-2005
Posted on Thursday, December 02, 2010 - 8:31 am:   

Dear Sirs :
I work during years in NIR and always I have a doubt about the correctly validation of a Model. I know is necesary have test samples, don�t outliers, measuring in the same conditions, etc. The real question is wich value calculate for to say this Model is validate. ASTM E 1655 say use minimum 20 samples and calculate T, if the T calculate is below a T table number with 95% confidence the Model is validate.
The question is: If you have 20 test samples and the result NIR and LAB for that, how you can say the Model is validate. ( sample cover all the range of the model). For me R2, sec, sep, etc ara not a good procedure for decide if a model is validate or not.
Some recomendations or comment.
Thans
Fernando Morgado

Add Your Message Here
Posting is currently disabled in this topic. Contact your discussion moderator for more information.