c# - How do I normalize Non-CSV data in Encog -
so i'm using jeff heaton's neural network library.
when trying solve iris plant classification problem have issue data normalization.
i able normalize csv file using following method:
public void normalizefile(fileinfo sourcedatafile, fileinfo normalizeddatafile, fileinfo normalizationconfigfile) { var wizard = new analystwizard(_analyst); wizard.wizard(sourcedatafile, _useheaders, analystfileformat.decpntcomma); var norm = new analystnormalizecsv(); norm.analyze(sourcedatafile, _useheaders, csvformat.english, _analyst); norm.produceoutputheaders = _useheaders; norm.normalize(normalizeddatafile); // save normalization configuration, can used later denormalize raw output. _analyst.save(normalizationconfigfile); }
so far good... program works high degree of accuracy.
the problem occurs when want enter values console application.
i have input data
- sepal width
- sepal length
- petal width
- petal length
each of these values has different high/low normalize these values can feed them network without writing csv file disk.
according link can using encog.util.arrayutil.normalizearray
:
i assume data stored in double[]
encog.util.arrayutil.normalizearray normalizer = new encog.util.arrayutil.normalizearray(); var normalizeddata = normalizer.process(datamatrix, 0, 1);//(yourdata, low, high)
Comments
Post a Comment