Creation of spiking neuron models applied in pattern recognition problems

Some spiking neuron models have proved to solve different linear and non-linear pattern recognition problems. Indeed, only one spiking neuron can generate comparable results as classical artificial neural network. However, depending on the classification problem, one spiking model could be better or less efficient than other. In this paper we propose a methodology to create spiking neuron models using Gene Expression Programming. The new models created are applied in eight pattern recognition problems. The results obtained are compared with previous results generated adopting the Izhikevich spiking neuron model. This first effort will help us to generate spiking neuron models which will be adaptable to a specific pattern recognition problem.