Chemometrics for Data Interpretation: Application of Principal Components Analysis (PCA) to Multivariate Spectroscopic Measurements

Extracting relevant and useful information from measurements is an issue of paramount importance and it can be considered as complementary to the process of data acquisition. This is a crucial point especially in the field of chemical measurements, where data sets can consist of hundreds or even thousands of variables so their interpretation can require long time. Chemometrics try to tackle this issue by applying mathematical and statistical tools to data coming from chemical, biological or medical analyses. Among possible methods, Principal Components Analysis (PCA) has found wide application in the I-M field thanks to its ability to identify patterns in acquired measurements and classify data in different groups. Possible applications span from chemicals detection [1] to concentration estimation of compounds in a given system [2]. Actually, many studies demonstrated the possibility to use PCA to process different kinds of data [3], in some cases coupling PCA to other tools such as artificial neural networks to improve the processing performance [4].