Just a basic question. Someone asked me to describe the algotrythm used to train the readout of an RNN, but I have no idea where to look. Could someone explain it to me? Thanks<
