Pattern Recognition of Vis/NIR Spectroscopy from White Vinegar Based on PLS and BP-ANN Model

Author

Wang, Li ; He, Yong ; Liu, Fei

Author_Institution

Zhejiang Univ., Hangzhou

fYear

2007

fDate

1-3 May 2007

Firstpage

1

Lastpage

4

Abstract

Visible/near infrared (Vis/NIR) transmittance spectroscopy was employed in discriminating white vinegar varieties. White vinegar samples were scanned in the Vis/NIR monochromatic instrument in transmission. It could get the scores of principal components (PCs) from original spectrum using partial least squares (PLS), the first 6 PCs picked according to the cumulative contribution rates would be taken as the inputs of back-propagation artificial neural network (BP-ANN), 240 samples from three varieties were used to build the model. Then this model was used to predict the varieties of 80 unknown samples and 97.5% resolving capability was achieved. It is indicated that Vis/NIR transmittance spectroscopy combined with PLS and BP-ANN is an effective measure to discriminate varieties of white vinegar.

Keywords

backpropagation; infrared spectroscopy; neural nets; pattern recognition; visible spectroscopy; BP-ANN model; PLS model; Vis/NIR spectroscopy; backpropagation artificial neural network; partial least squares; pattern recognition; visible/near infrared transmittance spectroscopy; white vinegar; Artificial neural networks; Information analysis; Infrared spectra; Least squares methods; Matrix decomposition; Pattern analysis; Pattern recognition; Personal communication networks; Principal component analysis; Spectroscopy; BP-ANN; PLS; Vis/NIR spectroscopy; pattern recognition; white vinegar;

fLanguage

English

Publisher

ieee

Conference_Titel

Instrumentation and Measurement Technology Conference Proceedings, 2007. IMTC 2007. IEEE

Conference_Location

Warsaw

ISSN

1091-5281

Print_ISBN

1-4244-0588-2

Type

conf

DOI

10.1109/IMTC.2007.379352

Filename

4258435