Interface | Description |
---|---|
Normalizer |
Interface for data set normalization methods.
|
Class | Description |
---|---|
DecimalScaleNormalizer |
Decimal scaling normalization method, which normalize data by moving decimal
point in regard to max element in training set (by columns) Normalization is
done according to formula: normalizedVector[i] = vector[i] / scaleFactor[i]
|
MaxMinNormalizer |
MaxMin normalization method, which normalize data in regard to min and max
elements in training set (by columns) Normalization is done according to
formula: normalizedVector[i] = (vector[i] - min[i]) / (max[i] - min[i])
This class works fine if max and min are both positive and we want to
normalize to [0,1]
|
MaxNormalizer |
Max normalization method, which normalize data in regard to max element in training set (by columns)
Normalization is done according to formula:
normalizedVector[i] = vector[i] / abs(max[i])
|
RangeNormalizer |
Performs normalization of a data set inputs and outputs to specified range.
|
ZeroMeanNormalizer |
Normalizes data sets by shifting all values in such way that data set has mean of 0 and standard deviation 1.
|
Copyright © 2019 Neuroph Project. All rights reserved.