A Matlab package for training and testing recurrent neural networks (RNN). The package also contains a few helper functions to assist with various training methods and tasks.
There main files are:
RNN.m: Class definition for the RNN. Contains methods for instantiating, training, and running the model. Training is accomplished mainly using the innate learning rule. For more information see: Laje, R., and Buonomano, D.V. (2013). Robust timing and motor patterns by taming chaos in recurrent neural networks. Nat. Neurosci. 16, 925�933. Hardy, N.F., Goudar, V., Romero-Sosa, J.L., and Buonomano, D. (2017). A Model of Temporal Scaling Correctly Predicts that Weber's Law is Speed-dependent. BioRxiv 159590.
TrainRNN_TempInv.m: Function to train RNNs to produce temporally invariant activity (temporal scaling).
TestRNN.m: Function to test RNNs.
TrainOut.m: Function to train the output units of the RNN.