Calibration Transfer Based on Affine Invariance for NIR without Transfer Standards

Abstract

Calibration transfer is an important field for near-infrared (NIR) spectroscopy in practical applications. However, most transfer methods are constructed with standard samples, which are expensive and difficult to obtain. Taking this problem into account, this paper proposes a calibration transfer method based on affine invariance without transfer standards (CTAI). Our method can be utilized to adjust the difference between two instruments by affine transformation. CTAI firstly establishes a partial least squares (PLS) model of the master instrument to obtain score matrices and predicted values of the two instruments, and then the regression coefficients between each of the score vectors and predicted values are computed for the master instrument and the slave instrument, respectively. Next, angles and biases are calculated between the regression coefficients of the master instrument and the corresponding regression coefficients of the slave instrument, respectively. Finally, by introducing affine transformation, new samples are predicted based on the obtained angles and biases. A comparative study between CTAI and the other five methods was conducted, and the performances of these algorithms were tested with two NIR spectral datasets. The obtained experimental results show clearly that, in general CTAI is more robust and can also achieve the best Root Mean Square Error of test sets (RMSEPs). In addition, the results of statistical difference with the Wilcoxon signed rank test show that CTAI is generally better than the others, and at least statistically the same.

Is Standardization Necessary for Sharing of a Large Mid-Infrared Soil Spectral Library?

“…For some properties such as clay and CaCO 3 , deployment of calibration transfer on the truly independent sets did not improve the performance even when the NSSC-KSSL library was used as a primary spectrometer. This is possibly due to the fact that the diversity of these soil properties were not well represented in either the transfer or the calibration set used to develop the predictive model [ 6 , 49 , 50 ]. For example, Ge et al [ 6 ] found improved performance of OC with PDS transfer spectra, but suggested that other soil properties might exhibit different variability in principal component space that could introduce differences in the prediction performance of the soil property under consideration with and without the calibration transfer.…”

“…In other words, despite all the potential additional variance introduced by use of a secondary spectrometer and PDS transformation, the large NSSC-KSSL library-based models often showed superior performance. Although some studies have shown that prediction using the PDS transfer spectra can be better than those of a direct calibration [ 28 ], others have indicated that the direct calibration generally outperforms the prediction using the PDS transfer spectra [ 50 ]. This could be attributed to the quality of the spectrometer with the PDS outperforming a direct calibration when the signal on one spectrometer is standardized to that of a higher quality spectrometer.…”

Advances in Near-Infrared Spectroscopy and Related Computational Methods

“…The Special Issue has met a remarkably positive feedback with many contributions submitted by numerous scholars and professional spectroscopists performing their active research in academia and industry, resulting in a collection of 30 publications including two exhaustive review articles [7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36]. The diversity in the application field has been well represented by the submitted manuscripts.…”

“…State-of-the-art analytical spectroscopy is based on sophisticated data-analytical methods. Development of new methods is, therefore, essential and benefits multiple applications [10,11,13,14,15,16]. Several articles focused on this direction, and the importance of research and development of calibration transfer methods is well reflected in this Special Issue [13,16].…”

“…Development of new methods is, therefore, essential and benefits multiple applications [10,11,13,14,15,16]. Several articles focused on this direction, and the importance of research and development of calibration transfer methods is well reflected in this Special Issue [13,16]. Interestingly, Beganović et al demonstrated that there exists room for improvement in fundamental aspects of analytical spectroscopy such as wavenumber region selection for subsequent calibration [3].…”