Chemometric Analysis of NMR Spectra
Autor: | Søren Balling Engelsen, Rasmus Bro, Nanna Viereck, Parvaneh Ebrahimi |
---|---|
Rok vydání: | 2018 |
Předmět: |
Multivariate analysis
Data exploration business.industry Computer science 010401 analytical chemistry Univariate Multivariate calibration Pattern recognition 01 natural sciences Nmr data 0104 chemical sciences NMR spectra database Chemometrics 03 medical and health sciences 0302 clinical medicine 030221 ophthalmology & optometry Artificial intelligence business |
Zdroj: | Modern Magnetic Resonance ISBN: 9783319282756 |
DOI: | 10.1007/978-3-319-28388-3_20 |
Popis: | NMR is one of the most powerful analytical techniques of our time. It allows detailed investigation of qualitative and quantitative characteristics of complex chemical and biological samples. The resulting NMR data provides a wealth of information about the samples, but the NMR data analysis has been and still is suffering from oversimplified approaches making it difficult to extract all the information efficiently. For instance, univariate methods that just use one or a few selected variables for the analysis from a whole spectrum lead to a huge loss of information. Such a simplifying approach reduces the chance of discovering new findings and truly learning about complex aspects of the samples investigated. Multivariate data analysis techniques allows for truly exploratory and comprehensive analysis of NMR data. This is particularly advantageous in the investigation of complex biological samples. Chemometrics can be helpful here by providing tools for unsupervised and supervised data exploration, multivariate calibration, classification and discrimination. This chapter presents some important steps in the pre-processing of NMR data as well as some of the most common chemometric techniques for data exploration and analysis. An example of NMR spectra of apple juice samples is given, to illustrate the power of the combination of NMR data and chemometrics. (Less) |
Databáze: | OpenAIRE |
Externí odkaz: |