Improving odor classification through self-organized lateral inhibition in a spiking olfaction-inspired network

In this study, we propose unsupervised learning of the lateral inhibition structure through inhibitory spike-timing dependent plasticity (iSTDP) in a computational model for multivariate data processing inspired by the honeybee antennal lobe. After exposing the network to a sufficient number of input samples, the inhibitory connectivity self-organizes to reflect the correlation between input channels. We show that this biologically realistic, local learning rule produces an inhibitory connectivity that effectively reduces channel correlation and yields superior network performance in a multivariate scent recognition scenario. The proposed network is suited as a preprocessing stage for spiking data processing systems, like for example neuromorphic hardware or neuronal interfaces.