基于主成分分析和模糊识别方法的生物分子太赫兹光谱识别

提出了一种基于主成分分析（PCA）和模糊模式识别方法的生物分子太赫兹（THz）光谱识别方法，并采用多种典型糖类和氨基酸生物分子的太赫兹透射光谱作为实验介质证明所提方法的可行性和有效性。首先，运用PCA方法对生物分子太赫兹光谱数据做降维处理，提取样品太赫兹光谱特征信息；然后，用获得的主成分得分矩阵代替原始太赫兹光谱数据输入到模糊模式识别分析模型中，运用基于择近原则的模糊模式识别方法对待定样品进行分类识别。实验结果表明以生物分子的太赫兹光谱作为数据特征，采用PCA与模糊识别相结合的方法实现生物分子的检测和识别是可行的，该方法为太赫兹光谱技术用于生物分子的鉴定和识别提供了一种新的有效的分析方法。

Terahertz spectra identification of biomolecules based on principal component analysis and fuzzy recognition

An identification method of terahertz (THz) spectra of biomolecules is proposed based on principal component analysis (PCA) and fuzzy pattern recognition,and THz spectra of several typical saccharide and amino acid biomolecules are used to verify its effectiveness and feasibility.PCA is used to decrease the dimensionality of original THz spectra variables and extract data features.Instead of the initial THz spectra variables,the selected principal component score matrix is input into the model of fuzzy pattern recognition,and fuzzy recognition based on principle of fuzzy closeness optimization is employed successfully to identify these samples.Results show that using THz spectroscopy of biological molecules as data feature,it is feasible to realize the detection and identification of biological molecules by the combination of PCA and fuzzy recognition.It provides a new and effective analytical method for the identification and recognition of biomolecules by using THz spectroscopy.