Building a Recommendation Engine: The XELOPES Library - Realtime Data Mining - page 256

inputMetaData.getMiningAttribute( modelTargetAttribute.
getName() );
// Classification:
int nall ¼ 0;
int nwrong ¼ 0;
while ( inputData.next() ) {

// Make prediction:
MiningVector vector ¼ inputData.read();
double predicted ¼ model.applyModelFunction(vector);
Category predCateg ¼ modelTargetAttribute.getCategory
(predicted);
double real ¼ vector.getValue(inputTargetAttribute);
Category realCateg ¼ inputTargetAttribute.getCategory
(real);

// Compare;
if (! predCateg.equals(tarCateg) )
nwrong ¼ nwrong + 1;
nall ¼ nall + 1;
};

¼

System.out.println("classification rate " + (100.0 -

wrong*100.0/nall) ); ■

12.1.4 The Mathematics Package

XELOPES is equipped with a useful package Math of basic mathematical
operations. It is only a utility package for the XELOPES analysis algorithms
and does not claim to be a complete mathematical library. Nevertheless, espe-
cially for numerical linear algebra, it contains some remarkable implementations
of general interest.
The Math package contains the following subpackages:

• Algebra : vectors, matrices, tensors, and important solver and factorizations
• Analysis : functions, derivatives, etc.
• Approximation: basic approximation methods
• Optimization: basic optimization methods
• Tools: useful tools

In the following, we want to delve into the package Algebra . It has three
subpackages: