header image
Home arrow Tutorial arrow Support Vector Machines and interpretation
Support Vector Machines and interpretation PDF Print E-mail
Written by Ravi Mallela, Equibits   
SVM’s are a relatively new technique that is being applied to QPSR Modeling. One of the major criticisms of SVM’s have been that they are “black boxes” that provide uninterpretable results. In reality, SVM’s lend themselves to easily interpretable models and results.
In this paper, we will demonstrate that non linear SVM’s, can provide high model accuracy with model information that can used to interpret results. A data set that contains mutagencity data will be used as a case study.
< Prev   Next >
Search website

Editorial flash
Prof. Svante Wold's 2003 New Year Message
Towards the end of the year we tend to reflect on what happened and what we accomplished in the year that passed. I am very happy that chemometrics goes forward rapidly on two fronts - exciting new applications, and better theory and methods.
Read more...
News flash
Tutorials and Editorials

The Tutorial and Editorial sections have now been updated with a link to their respective pdf files for you to download.

Regards,
Johan T
Tutorial flash
Prediction in linear models
Chemometrics uses an empirical approach to modelling of data. One can say that data is used to generate the model. This is the main criticism from the community of statisticians. It is in the nature of natural sciences to argue for a mathematical model, and use data to estimate unknown parameters in the model.
Read more...