Abstract
A discriminant analysis technique using wavelet transformation (WT) and influence matrix analysis (CAIMAN) method is proposed for the near infrared (NIR) spectroscopy classification. In the proposed methodology, NIR spectra are decomposed by WT for data compression and a forward feature selection is further employed to extract the relevant information from the wavelet coefficients, reducing both classification errors and model complexity. A discriminant-CAIMAN (D-CAIMAN) method is utilized to build the classification model in wavelet domain on the basis of reduced wavelet coefficients of spectral variables. NIR spectra data set of 265 salviae miltiorrhizae radix samples from 9 different geographical origins is used as an example to test the classification performance of the algorithm. For a comparison, k-nearest neighbor (KNN), linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA) methods are also employed. D-CAIMAN with wavelet-based feature selection (WD-CAIMAN) method shows the best performance, achieving the total classification rate of 100% in both cross-validation set and prediction set. It is worth noting that the WD-CAIMAN classifier also shows improved sensitivity, selectivity and model interpretability in the classifications.
Highlights
Near infrared (NIR) spectroscopy has the advantage of being fast, robust, nondestructive and especially suitable for the online application
We introduced a discriminant classication method named discriminant-Classication and in°uence matrix analysis (CAIMAN) (D-CAIMAN) to simultaneously classify the NIR spectral data of samples
A new algorithm, WD-CAIMAN, was proposed for NIR spectra classication based on the waveletbased feature selection and D-CAIMAN discrimination methods
Summary
Near infrared (NIR) spectroscopy has the advantage of being fast, robust, nondestructive and especially suitable for the online application. With the development of modern instruments and chemometrics, NIR spectroscopy has been widely applied for the quantitative and qualitative analysis in large areas, such as agriculture, pharmaceuticals, food, textiles and polymer production.[1] the physical and chemical information cannot always be extracted straightforwardly from the spectra due to the existence of band overlapping, multicollinearity, poor signal-to-noise ratio, baseline °uctuations, and so on. In order to overcome these di±culties in NIR spectral analysis, chemometrics has to be used for preprocessing, modeling, validation, etc. A main part of chemometrics is multivariate data analysis, which is essential for qualitative and quantitative assays based on NIR spectroscopy. Statistical classication with NIR data has been used in a number of scientic publications and practical applications.[1,2]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
