Novel nonlinear multivariate calibration methods
1998 (English)Doctoral thesis, comprehensive summary (Other scientific)
The interest fornonlinear modeling has increasedconcomitantly with the increased awareness of the ubiquity ofnonlinear systems. The traditional solution in analyticalchemistry, to linearize the system by some sort of samplepretreatment, is not always an option in process analyticalchemistry. Sample pretreatment merely introduces unwantedsources of error. The analysis should be performed in-line, orat least on-line, and facilitate a rapid prediction of keyprocess parameters.
Keywords: multivariate calibration, nonlinear modeling,PLS, neural networks, linearing transform
Place, publisher, year, edition, pages
Stockholm: Kemi , 1998. , 50 p.
IdentifiersURN: urn:nbn:se:kth:diva-2649ISBN: 91-7170-266-0OAI: oai:DiVA.org:kth-2649DiVA: diva2:8316
NR 201408052000-01-012000-01-01Bibliographically approved