Abstract-Application of terahertz spectroscopy and chemometrics for discrimination of transgenic camellia oil

Jianjun Liu, Lanlan Fan, Yuanming Liu, Lili Mao, Jianquan Kan,

PLS-WLDA, an alternative method, is performed on terahertz spectra data. The identification result of PLS-WLDA of terahertz spectra using different preprocessing type (raw, offset and derivative), it is possible to correctly identify 97% of all samples set. Compared with the SPA-WLDA, PLS-WLDA has the better discrimination performance. However, in the aspect of robustness, SPA–WLDA is better than PLS-WLDA.

https://www.sciencedirect.com/science/article/pii/S1386142518307753

Discrimination of transgenic edible oil has become the focus of attention in the field of food safety. In this paper, we propose a method for discrimination of transgenic edible oils by using terahertz spectroscopy combine with weighted linear discriminant analysis (WLDA). To evaluate the lustiness of the model, we employ successive projection arithmetic (SPA) and partial least squares (PLS) to verify the discrimination performance through variable selection. The results demonstrate that the SPA-WLDA model has higher classification accuracy than PLS-WLDA. In conclusion, terahertz spectroscopy is coupled with chemometrics is an effective method for discriminating various types of transgenic edible oils.