A Neural Assembly-Based State Machine Design

Neural Assembly Computing (NAC) is a framework that employs spiking neural networks to model the computational processes of biological neural cell assemblies. In this paper, we propose a method to design state machines based on NAC, capable of generating complex and intelligent behaviors. This study, as an example, presents a simple algorithm for recognizing a sequence of distinct symbols (geometric shapes) by utilizing four states that mirror the activity of neural assemblies. Through simulation outcomes, we demonstrate that NAC provides a biologically realistic and efficient strategy for implementing state machines and algorithms.