Load the training data.

X = digitsmall_dataset;

The training data is a 1-by-500 cell array, where each cell containing a 28-by-28 matrix representing a synthetic image of a handwritten digit.

Train an autoencoder with a hidden layer containing 25 neurons.

hiddenSize = 25;
autoenc = trainAutoencoder(X,hiddenSize,...
        'L2WeightRegularization',0.004,...
        'SparsityRegularization',4,...
        'SparsityProportion',0.15);

Create a Simulink model for the autoencoder, autoenc.

generateSimulink(autoenc)