de.aitools.ir.retrievalmodels.representer
Class LatentSemanticIndexing.QualityReductionStrategyImpl

java.lang.Object
  de.aitools.ir.retrievalmodels.representer.LatentSemanticIndexing.QualityReductionStrategyImpl

Enclosing class:

LatentSemanticIndexing

public class LatentSemanticIndexing.QualityReductionStrategyImpl
extends java.lang.Object

Reduces the matrix dynamically to the lowest rank that satisfies the given quality. The threshold has to be between 0.0 and 1.0 and resembles the degree of approximation of the original matrix in percentage. For calculation, we use the ratio between the Frobenius norm of the reduced matrix and the original matrix.

Constructor Summary

LatentSemanticIndexing.QualityReductionStrategyImpl(double quality)

Method Summary

int

reduce()

Methods inherited from class java.lang.Object

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

Constructor Detail

LatentSemanticIndexing.QualityReductionStrategyImpl

public LatentSemanticIndexing.QualityReductionStrategyImpl(double quality)

Parameters:

quality - specifies the quality, by which the original matrix will be approximated

Method Detail

reduce

public int reduce()