Echo spiking neural P systems

摘要

Nonlinear spiking neural P (NSNP) systems are distributed parallel neural-like computing models that abstract the nonlinear spiking mechanisms of biological neurons. Echo state network (ESN) is a new type of recurrent neural network (RNN) that can overcome the disadvantages of traditional RNN. Inspired by the structure of ESN, this study proposes a new variant of NSNP systems, called echo spiking neural P (ESNP) systems, or the ESNP model. An ESNP system is essentially a specialized NSNP system equipped with an input layer and a readout layer. The ESNP system can also be viewed as a variant of ESN, where a specialized NSNP system is used as its reservoir. Owing to the use of spiking neurons, the state equation of ESNP model is different from the state equation of the ESN model. The proposed ESNP model is a recurrent-like model, and we used time-series forecasting as a case study to prove its capabilities. Experimental results demonstrate the effectiveness of the proposed ESNP model for time-series forecasting.