Explanation of Hidden and Visible in a layer.

mnnlConfig

 MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
           .seed(seed)
           .gradientNormalization(GradientNormalization.ClipElementWiseAbsoluteValue)
           .gradientNormalizationThreshold(1.0)
           .iterations(iterations)
           .momentum(0.5)
           .momentumAfter(Collections.singletonMap(3, 0.9))
           .optimizationAlgo(OptimizationAlgorithm.CONJUGATE_GRADIENT)
           .list(4)
           .layer(0, new RBM.Builder().nIn(numRows*numColumns).nOut(500)
                         .weightInit(WeightInit.XAVIER).lossFunction(LossFunction.RMSE_XENT)
                         .visibleUnit(RBM.VisibleUnit.BINARY)
                         .hiddenUnit(RBM.HiddenUnit.BINARY)
                         .activation("relu")
                         .build())