Spiking neural networks, thanks to their sensitivity to the timing of the inputs, are a promising tool for unsupervised processing of spatio-temporal data. However, they do not perform as well as the traditional machine learning approaches and their real-world applications are still limited. Various supervised and reinforcement learning methods for optimising spiking neural networks have been proposed, but more recently the evolutionary approach regained attention as a tool for training neural networks. Here, we describe a simple evolutionary approach for optimising spiking neural networks. This is the first published use of evolutionary algorithm to develop hyperparameters for fully unsupervised spike-timing-dependent learning for pattern clustering using spiking neural networks. Our results show that combining evolution and unsupervised learning leads to faster convergence on the optimal solutions, better stability of fit solutions and higher fitness of the whole population than using each approach separately.

Abstract concepts are rules about relationships such as identity or sameness. Instead of learning that specific objects belong to specific categories, the abstract concept of same/different applies to any objects that an organism might encounter, even if those objects have never been seen before. In this paper we investigate learning of abstract concepts by computer, in order to recognize same/different in novel data never seen before. To do so, we integrate recursive self-organizing maps with the data they are processing into a single graph to enable a brain-like self- adaptive learning system. We perform experiments on simple same/different datasets designed to resemble those used in animal experiments and then show an example of a practical application of same/different learning using the approach.