A spike trains encoding and decoding solution for the spiking neural networks

Abstract

This paper proposes a spike train encoding and decoding solution to process input and output signals for the spiking neural networks. The efficiency of the proposed solution is verified by the experimental tasks: The XOR classification problem and the aerodynamic coefficients identification of an aircraft from the data sets are recorded from flights. The results show that the proposed encoding and decoding solution has a higher convergence rate to the set values, and the mean squared error smaller than another solution is introduced in this research.