A Spiking Neuron Implementation of Genetic Algorithms for Optimization.

We designed freely scalable ensembles of spiking neurons to carry out the operations required to run a genetic algorithm, thereby opening up possibilities for making use of efficient neuromorphic hardware. Two types of implementation are explored that offer a complexity trade-off between computational space and time, with both designs having linear energy complexity. The designs were implemented in a simulator to successfully solve the one-max optimization problem, serving as a proof of concept for running genetic algorithms as spiking neural networks.