PLSFilter

Package

weka.filters.supervised.attribute

Synopsis

Runs Partial Least Square Regression over the given instances and computes the resulting beta matrix for prediction.
By default it replaces missing values and centers the data.

For more information see:

Tormod Naes, Tomas Isaksson, Tom Fearn, Tony Davies (2002). A User Friendly Guide to Multivariate Calibration and Classification. NIR Publications.

StatSoft, Inc.. Partial Least Squares (PLS).

Bent Jorgensen, Yuri Goegebeur. Module 7: Partial least squares regression I.

S. de Jong (1993). SIMPLS: an alternative approach to partial least squares regression. Chemometrics and Intelligent Laboratory Systems. 18:251-263.

Options

The table below describes the options available for PLSFilter.

Option	Description
algorithm	Sets the type of algorithm to use.
debug	Turns on output of debugging information.
numComponents	The number of components to compute.
performPrediction	Whether to update the class attribute with the predicted value.
preprocessing	Sets the type of preprocessing to use.
replaceMissing	Whether to replace missing values.

Capabilities

The table below describes the capabilites of PLSFilter.

Capability	Supported
Class	Date class, Numeric class
Attributes	Missing values, Date attributes, Numeric attributes
Min # of instances	0