Temporal pattern classification in the model of olfactory cortex

In this paper, we present a neural network model which has the architecture of the olfactory cortex. Although the olfactory cortex is considered to process the spatio-temporal information generated in the olfactory bulb, it is not clear what kind of information is brought to the cortex and what kind of pattern is being generated in the cortex by the input temporal pattern. It is shown that certain types of the input pulse train induce a long lasting oscillation spreading all over the network. The oscillation is organized based on the Hebbian learning rule so as to produce a characteristic pattern for each input pattern. The patterns of spatio-temporal oscillation become different depending on input patterns. The patterns are also changed depending on the learning processes. It is highly possible that this type of spatio-temporal patterns is essential for the odor recognition.