How to do denormalization?

DL4J
1

Hello! Help me please!

I have a recurrent neural network for regression over 100 time steps.
One time step has the initial form:
[98.99920226267315, 98.74175067082459, 99.00645442019002, 98.69461164696499, 183762.0]

In order to enter data into the model, the data undergoes a normalization process

DataNormalization normalizer = new NormalizerStandardize();
normalizer.fit(trainData);
...
finalData.setPreProcessor(normalizer);

After training the model, I input finalData and get the next time step in the output. But the data looks like:

[0.6522, 1.1029, 1.0113, 1.2451, 5.5019]

How can I denormalize the data and compare it to the original values?

2

@st331man use the revertFeatures/revertLabels in the class.

3

@agibsonccc Thank you, it worked.