Learning Classification in the Olfactory System of Insects

Abstract

We propose a theoretical framework for odor classification in the olfactory system of insects. The classification task is accomplished in two steps. The first is a transformation from the antennal lobe to the intrinsic Kenyon cells in the mushroom body. This transformation into a higher-dimensional space is an injective function and can be implemented without any type of learning at the synaptic connections. In the second step, the encoded odors in the intrinsic Kenyon cells are linearly classified in the mushroom body lobes. The neurons that perform this linear classification are equivalent to hyperplanes whose connections are tuned by local Hebbian learning and by competition due to mutual inhibition. We calculate the range of values of activity and size of the network required to achieve efficient classification within this scheme in insect olfaction. We are able to demonstrate that biologically plausible control mechanisms can accomplish efficient classification of odors.

Self-organization in the olfactory system: one shot odor recognition in insects

“…Several attempts have been made to understand odor discrimination in models of the olfactory system [17,51,13,26,9,3,32,19]. The model discussed here is, however, fundamentally different in providing a classification scheme that solely relies on the fan-in, fan-out properties of synaptic connections, the known locus of learning in the MB and otherwise entirely non-specific connectivity.…”

“…While the current study is closer to the biological systems than our previous work [19], it is not yet describing a specific insect and specializing to a neuron model designed to match honeybee KCs would be pretentious. Nevertheless, the chosen model is as a type 1 model well suited to allow low firing frequencies and the observed sparse activity in KCs [35].…”

“…The general statistical properties of non-specific connectivities restrict the activity levels for successful classification to a small set of allowed levels [19], and the olfactory system, at least in locust, indeed seems to have a gain control mechanisms to regulate activity levels in the in the AL and the MB [43]. The gain control within the AL is mediated by inhibitory interneurons whereas the iKCs of the MB are subject to a feed-forward periodic inhibition.…”

“…It has become clear in numerous works [15,14,27,41,3] that systems built with more realistic models of biological components can be quantitatively and qualitatively different from the more abstract connectionist approaches. In the context of this article, for example, non-disjoint representations of odors that were easily implemented in our previous more abstract work on the olfactory system of insects [19] could not be observed under any conditions in the more realistic system described here.…”

A modelling framework for the olfactory system of the honeybee using GeNN (GPU enhanced Neuronal Network simulation environment)

“…Odour mixtures are represented as a non-trivial combination of constituent odours' representations, and formation of long-term memories associated with such odour mixtures has been shown to induce volume changes in glomeruli indicating a cross-inhibitory effect between neural codings [2]. The modelling will also consider how mechanisms might implement known classification rules, such as in the models of insect olfactory classification by Huerta et al [3] and Nowotny et al [4].…”