use MultiLayerConfiguration.Builder.setInputType(InputType) with InputType.convolutional(height,width,depth), for CNN data with
shape [minibatchSize,depth,height,width]. For image data that has been flattened into a row vector per example
(shape [minibatchSize,depth*height*width]) instead use InputType.convolutionalFlat(height,width,depth)
use MultiLayerConfiguration.Builder.setInputType(InputType) with InputType.convolutional(height,width,depth), for CNN data with
shape [minibatchSize,depth,height,width]. For image data that has been flattened into a row vector per example
(shape [minibatchSize,depth*height*width]) instead use InputType.convolutionalFlat(height,width,depth)
As of 0.7.3 - Feb 2017. Not used; neural network params are initialized by the parameter initializaters.
Furthermore, most implementations are unsupported or no-op.
Use MultiLayerConfiguration.Builder.setInputType(InputType) to set nIns
and add preprocessors as required. This can be done using builder.setInputType(InputType.convolutional(height, width, channels))
For image data that has been flattened into a row vector per example (shape [minibatchSize,depth*height*width])
instead use InputType.convolutionalFlat(height,width,depth).