A Survey of Learning Spiking Neural P Systems and A Novel Instance
Yunhui Chen, Ying Chen, Gexiang Zhang, Prithwineel Paul, Tianbao Wu, Xihai Zhang, Haina Rong and Xiaomin Ma
In the last few decades membrane computing has established itself as an important branch of natural computing. Investigating computational power, complexity aspects and real-world applications of different variants of membrane computing models have been a successful direction of research. In recent years with the invention of efficient learning algorithms, many researchers have concentrated their research into construction of intelligent biological computing systems inspired by the working of neurons in human brains to emulate human thinking. Spiking neural P systems are such types of computing systems. In this paper we survey spiking neural P systems (i.e., neural-like membrane computing models) with learning ability, their architecture, learning mechanism and compare these models, discuss their advantages and disadvantages and application of these models in solving real-world problems. We further discuss the learning mechanism of associative memory network based on spiking neural P systems with white holes and weights. At the end, we discuss some new ideas to further extend the study of membrane computing models having learning ability.
Keywords: Membrane computing, spiking neural networks, spiking neural P systems, neural plasticity, structural plasticity, machine learning