Support vector machines are used in data mining (predictive modeling, to be specific) for classification of records, by learning from training data. Support vector machines use decision surfaces that separate records. They rely on optimization techniques to maximize separate margins between classes, and kernel functions to accommodate non-linearity. Model parameters can be tuned on the basis of model performance with hold-out samples.
The Institute for Statistics Education offers an extensive glossary of statistical terms, available to all for reference and research. To celebrate the International Year of Statistics in 2013, we will provide a statistical term every week, delivered directly to your inbox. Make it your New Year's resolution to improve your own statistical knowledge! Sign up here.
Rather not have more email? Bookmark our "Stats Word of the Week" page.