偏最小二乘回归
化学计量学
灵芝
灵芝
孢子
线性判别分析
近红外光谱
食品科学
淀粉
化学
材料科学
色谱法
数学
植物
生物
统计
神经科学
作者
Xinyu Shi,Xiuqi Gan,Xinbo Wang,Jialin Peng,Zhenhao Li,Xue‐Qian Wu,Qingsong Shao,Ailian Zhang
标识
DOI:10.1016/j.lwt.2022.113829
摘要
As a valuable traditional Chinese medicine, Ganoderma lucidum spore powder is often mixed with dyed starch to reduce production costs. In this report, we developed a near-infrared (NIR) spectroscopy method to better identify the authenticity of spore powder samples and predict their adulteration content. The model based on NIR spectral data of samples combined with partial least squares discriminant analysis rapidly and accurately identified the adulterated G. lucidum spore powder. Partial least square (PLS) and artificial neural network models were used for regression analysis to predict the level of adulteration. The results showed that the PLS model was better in determining the adulteration content of G. lucidum spore powder, with correlation coefficients in calibration and validation greater than 0.99, and the root mean square errors of cross-validation and prediction were 0.9218% and 3.1127%, respectively. Thus, combining NIR spectroscopy and chemometric methods is an effective and rapid approach to identifying and quantifying the adulteration level of G. lucidum spore powder.
科研通智能强力驱动
Strongly Powered by AbleSci AI