Shuffling Recurrent Neural Networks

We propose a novel recurrent neural network model, where the hidden state ht is obtained by permuting the vector elements of the previous hidden state ht-1 and adding the output of a learned function b(xt) of the input xt at time t. In our model, the prediction is given by a second learned function, which is applied to the hidden state s(ht). The method is easy to implement, extremely efficient, and does not suffer from vanishing nor exploding gradients. In an extensive set of experiments, the method shows competitive results, in comparison to the leading literature baselines.