闲鹤野云
第2楼2011/11/24
Principal Component Analysis:
PCA is a method for reducing the 100 variables (wavelength data) in each spectrum down to just a few important variables. These variables are often referred to as latent variables, principal components, factors, eigenvectors, etc, and are vectors. This manual will refer to them as PC’s. The dot product of these vectors with the spectral data yields scalars called “PC scores”. Unknowns can be identified by comparing the PC scores of unknown materials to those of the model.
比如这个方法,也叫主成分分析法(PCA)
闲鹤野云
第3楼2011/11/24
Spectral Matching:
As an alternative to PCA, Spectral Matching may be used as a material identification method. This is particularly useful for large numbers of categories. Spectral Matching compares the shape of each spectrum with each spectrum in the library and assigns a “degree of match” value ranging from ‑1 (perfectly anti-matched) to +1 (perfect match) using a proprietary algorithm. The library entries that have the highest match values to the unknown sample are then used to identify the unknown.
这也是一种定性分析方法,叫做光谱比对法,这主成分分析的中替代方法,用于材料定性分析。通过与已知材料的光谱信息库进行比较,根据光谱相似度(负一相当于无相似性,到+1=全部拟合),来判断未知材料可能是什么。一般简写为SM法。