The SDL Component Suite is an industry leading collection of components supporting scientific and engineering computing. Please visit the SDL Web site for more information....


Class: TKohonen
Declaration: procedure StandardizeData;

The method StandardizeData scales the training data in a way that the variables (=columns of TrainData) have zero mean and unit standard deviation. Whether or not to use standardization depends on the data; in many cases standardization of the training data improves the clustering results.

Hint 1: Please note that the method StandardizeData changes the training data and stores the transformation equations in the network. Thus you must not apply StandardizeData twice to the same data set.

Hint 2: Background information on the standardization of data can be found in Fundamentals of Statistics.

Example: This method is used in the following example program (see for downloading the code): kohmap

Last Update: 2012-Okt-20