org.knime.base.node.mine.bfn.fuzzy
Class FuzzyBasisFunctionFactory

java.lang.Object
  org.knime.base.node.mine.bfn.BasisFunctionFactory
      org.knime.base.node.mine.bfn.fuzzy.FuzzyBasisFunctionFactory

public class FuzzyBasisFunctionFactory
extends BasisFunctionFactory

Basic interface for all basis function algorithms. Provides the function getNewBasisFunction() to initialise a new prototype. This interface is needed in order to create new prototypes in the general BasisFunctionLearner. Hence a BasisFunctionLearner would be initialised with an object of type BasisFunctionFactory. It is used as factory to create basisfunctions. One implementation of the BasisFunctionFactory; here represents the FuzzyBasisFunctionFactory object.

Author:

Thomas Gabriel, University of Konstanz

See Also:

FuzzyBasisFunctionLearnerRow, commit(RowKey, DataCell, DataRow)

Field Summary

Fields inherited from class org.knime.base.node.mine.bfn.BasisFunctionFactory

CLASS_COLUMN

Constructor Summary

FuzzyBasisFunctionFactory(int norm, int shrink, DataTableSpec spec, String[] targetColumns, int distance)
Creates a new factory fuzzy basisfunction along with a Norm and a Shrink function.

Method Summary

BasisFunctionLearnerRow	commit(RowKey key, DataCell classInfo, DataRow row) Creates and returns a new row initialised with a class label and a center vector.
int	getNorm() Returns the upper bound for conflicting instances.
int	getShrink() Returns the lower bound for non-conflicting instances.
void	save(ModelContent pp) Saves to model content.

Methods inherited from class org.knime.base.node.mine.bfn.BasisFunctionFactory

createModelSpec, findDataColumns, getDistance, getMaximums, getMinimums, getModelSpec

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

FuzzyBasisFunctionFactory

public FuzzyBasisFunctionFactory(int norm,
                                 int shrink,
                                 DataTableSpec spec,
                                 String[] targetColumns,
                                 int distance)

Creates a new factory fuzzy basisfunction along with a Norm and a Shrink function.

Parameters:

norm - the choice of fuzzy norm

shrink - the choice of shrink procedure

spec - the data to retrieve all columns and class info from

targetColumns - the class info column in the data

distance - the choice of distance function

Method Detail

commit

public BasisFunctionLearnerRow commit(RowKey key,
                                      DataCell classInfo,
                                      DataRow row)

Creates and returns a new row initialised with a class label and a center vector.

Specified by:

commit in class BasisFunctionFactory

Parameters:

key - the key for this row

row - the initial center vector

classInfo - the class info

Returns:

A new basisfunction

getNorm

public final int getNorm()

Returns the upper bound for conflicting instances.

Returns:

the upper bound for activation

getShrink

public final int getShrink()

Returns the lower bound for non-conflicting instances.

Returns:

the lower bound for activation

save

public void save(ModelContent pp)

Saves to model content.

Overrides:

save in class BasisFunctionFactory

Parameters:

pp - the model content this is saved to.