Multivariate statistical analysis of the spectroscopic data-sets – building open-source, user-friendly interface for data evaluation (Talk)

Abstract

The main goal of this project is to establish a reliable and robust model for fast identification and detection of counterfeited food products by using FTIR spectroscopy in conjunction with R Platform for statistical computing. FTIR preliminary analysis will be performed on samples which adulterations are quite often in recent years due to the high economic value and limited supply. Collecting the spectral data is important but only first step in the analysis of samples. Next, crucial for achieving the goal of the project, is data evaluation stage. Proposed in this contribution work-flow is employing open source R Environment for pre-processing and statistical analysis of the spectral data. In order to enable user-friendly interface Shiny package will be used. In this way power and flexibility of R is hidden for the end-user. Scalability of this solution - even by implementing Big Data algorithms gives high degree of freedom while Open Source license will help to disseminate the work-flow procedures to developing countries without extra costs, and additionally can be used to interpret straightforwardly analyzed samples even by non-specialized person.

Krzysztof Banas

Principal Research Fellow

I work as beam-line scientist at Singapore Synchrotron Light Source. My research interests include application of advanced statistical methods for hyperspectral data processing (dimension reduction, clustering and identification).