Quansheng Chen

Rapid measurement of total acid content (TAC) in vinegar using near infrared spectroscopy based on efficient variables selection algorithm and nonlinear regression tools.

Total acid content (TAC) is an important index in assessing vinegar quality. This work attempted to determine TAC in vinegar using near infrared spectroscopy. We systematically studied variable selection and nonlinear regression in calibrating regression models. First, the efficient spectra intervals were selected by synergy interval PLS (Si-PLS); then, two nonlinear regression tools, which were extreme learning machine (ELM) and back propagation artificial neural network (BP-ANN), were attempted. Experiments showed that the model based on ELM and Si-PLS (Si-ELM) was superior to others, and the optimum results were achieved as follows: the root mean square error of prediction (RMSEP) was 0.2486 g/100mL, and the correlation coefficient (R(p)) was 0.9712 in the prediction set. This work demonstrated that the TAC in vinegar could be rapidly measured by NIR spectroscopy and Si-ELM algorithm showed its superiority in model calibration.

Identification of green tea’s (Camellia sinensis (L.)) quality level according to measurement of main catechins and caffeine contents by HPLC and support vector classification pattern recognition

High performance liquid chromatography (HPLC) was identified green teas quality level by measurement of catechins and caffeine content. Four grades of roast green teas were attempted in this work. Five main catechins ((-)-epigallocatechin gallate (EGCG), (-)-epigallocatechin (EGC), (-)-epicatechin gallate (ECG), (-)-epicatechin (EC), and (+)-catechin (C)) and caffeine contents were measured simultaneously by HPLC. As a new chemical pattern recognition, support vector classification (SVC) was applied to develop identification model. Some parameters including regularization parameter (R) and kernel parameter (K) were optimized by the cross-validation. The optimal SVC model was achieved with R=20 and K=2. Identification rates were 95% in the training set and 90% in the prediction set, respectively. Finally, compared with other pattern recognition approaches, SVC algorithm shows its excellent performance in identification results. Overall results show that it is feasible to identify green teas quality level according to measurement of main catechins and caffeine contents by HPLC and SVC pattern recognition.

Determination of free amino acid content in Radix Pseudostellariae using near infrared (NIR) spectroscopy and different multivariate calibrations

Near infrared (NIR) spectroscopy combined with multivariate calibration was attempted to analyze free amino acid content of Radix Pseudostellariae. The original spectra of Pseudostellariae samples in wavelength range of 10000-4000 cm(-1) were acquired. Partial least squares (PLS), kernel PLS (k-PLS), back propagation neural network (BP-NN), and support vector regression (SVR) algorithms were performed comparatively to develop calibration models. Some parameters of the calibration models were optimized by cross-validation. The performance of BP-NN model was better than PLS, k-PLS, and SVR models. The root mean square error of prediction (RMSEP) and the correlation coefficient (R) of BP-NN model were 0.687 and 0.889 in prediction set respectively. Results showed that NIR spectroscopy combined with multivariate calibration has significant potential in quantitative analysis of free amino acid content in Radix Pseudostellariae.

Automated tea quality classification by hyperspectral imaging

A hyperspectral imaging technique was attempted to classify green tea. Five grades of green tea samples were attempted. A hyperspectral imaging system was developed for data acquisition of tea samples. Principal component analysis was performed on the hyperspectral data to determine three optimal band images. Texture analysis was conducted on each optimal band image to extract characteristic variables. A support vector machine (SVM) was used to construct the classification model. The classification rates were 98% and 95% in the training and prediction sets, respectively. The SVM algorithm shows excellent performance in classification results in contrast with other pattern recognitions classifiers. Overall results show that the hyperspectral imaging technique coupled with a SVM classifier can be efficiently utilized to classify green tea.

Qualitative identification of tea by near infrared spectroscopy based on soft independent modelling of class analogy pattern recognition

Near-infrared (NIR) spectroscopy has been successfully utilised for the rapid identification of tea varieties. The spectral features of each tea category are reasonably differentiated in the NIR region and the spectral differences provided enough qualitative spectral information for identification. Soft independent modelling of class analogy (SIMCA) as the pattern recognition was applied in this paper. In this study, both α-error (i.e. the rejection of correct samples from their class) and β-error (i.e. the acceptance of objects that do not belong to that class) are focused on. Four tea classes from Longjing tea, Biluochun tea, Qihong tea and Tieguanyin tea were modelled separately by principal component analysis (PCA). The results showed that at the 99% confidence level, the α-errors were equal to 0.1 only for the Longjing tea class when training and 0.2 only for the Biluochun tea class when testing, while the remaining α-errors and all β-errors were equal to zero. The study demonstrated that NIR spectroscopy technology with a SIMCA pattern recognition method can be successfully applied as a rapid method to identify the class of tea.