Studies on reservoir initialization and dynamics shaping in echo state networks

The fixed random connectivity of networks in reservoir com- puting leads to significant variation in performance. Only few problem specific optimization procedures are known to date. We study a general initialization method using permutation matrices and derive a new un- supervised learning rule based on intrinsic plasticity (IP) for echo state networks. Using three different benchmarks, we show that networks with permutation matrices for the reservoir connectivity have much longer mem- ory than the other methods, but are also able to perform highly non-linear mappings. We also show that IP based on sigmoid transfer functions is limited concerning the output distributions that can be achieved.