Classification Analysis of Spectral Data Using Chemometrics

by | Jul 17, 2018

This presentation will help you with classification and identification of spectral data to decide what technique to use whether it be wavelength correlation, Euclidean distance (ED), soft...

There are numerous methods for classification and identification of spectral data. This on-demand presentation by Frederick Long will help you to decide what technique to use for the problem at hand by offering several chemometric or statistical approaches.

There are two general categories of classification problems with the first most commonly used in many industries being the problem of chemical identification. There are several algorithmic approaches for this usually going by the name of wavelength correlation or the related method of Euclidean distance. The second problem associated with classification is often just referred to as classification and this is a more sensitive problem. There are numerous techniques that can be used on spectral data: principal component analysis (PCA), soft independent modeling of class analogies (SIMCA) and partial least squares discriminant analysis (PLS-DA), which is related to the method of PLS that is often used for quantitative modeling.

Learning outcomes
By viewing this presentation you will learn more about these methods and be provided with some simple examples to help your understanding. Whether it is a simple problem that can be resolved with simple spectral fingerprint matching or a problem that requires a more sensitive handling, there are algorithmic approaches in both cases that are very reliable.

Frederick LongFrederick Long founded Spectroscopic Solutions in 2001. Spectroscopic Solutions is a leading consulting and training firm in the areas of spectroscopy, chemometrics and method validation. Spectroscopic Solutions has done work for leading pharmaceutical, chemical, consumer health, and oil and gas companies. Dr. Long earned his Ph.D. in Chemical Physics from Columbia University and a S.B. and S.M. in Physics from MIT.
 

 

Related Content

Advertisment

Advertisment

Advertisement