Meaning of the net.parameter 'divideby : 255' with userguide example Neuralnet Classifier?

User 5217 | 5/22/2016, 4:38:02 PM

I can not find an explanation of parameter 'divideby' in the API documents.

Comments

User 1774 | 5/22/2016, 5:59:49 PM

Hi Emma, It's here: https://dato.com/products/create/docs/generated/graphlab.neuralnet_classifier.create.html There's no anchor so search for divideby in this page. The documentation says: The value by which to scale the input data before it is inserted into the network. When the data is pictures with the pixels ranging from 0 to 255, one of the best practices is to normalize it to a [0, 1) range. Another would be z-score normalization (remove the mean, then divide by stdev). But that's a different discussion altogether :)