Yavuz, Esin and Nowotny, Thomas (2016) Input-modulation as an alternative to conventional learning strategies. In: Proceedings of the ICANN 2016 Conference. Lecture Notes in Computer Science, 9886 . Springer, pp. 54-62. ISBN 0302-9743

PDF - Accepted Version Download (855kB)

Abstract

Animals use various strategies for learning stimulus-reward associations. Computational methods that mimic animal behaviour most commonly interpret learning as a high level phenomenon, in which the pairing of stimulus and reward leads to plastic changes in the final output layers where action selection takes place. Here, we present an alternative input-modulation strategy for forming simple stimulus-response associations based on reward. Our model is motivated by experimental evidence on modulation of early brain regions by reward signalling in the honeybee. The model can successfully discriminate dissimilar odours and generalise across similar odours, like bees do. In the most simplified connectionist description, the new input- modulation learning is shown to be asymptotically equivalent to the standard perceptron.

Item Type:	Book Section
Keywords:	Reinforcement learning, olfactory system, spiking neural network
Schools and Departments:	School of Engineering and Informatics > Informatics
Subjects:	Q Science > QA Mathematics > QA0299 Analysis. Including analytical methods connected with physical problems Q Science > QP Physiology > QP0351 Neurophysiology and neuropsychology > QP0361 Nervous system Q Science > QP Physiology > QP0351 Neurophysiology and neuropsychology > QP0431 Senses > QP0448 Special senses > QP0455 Chemical senses. Chemoreceptors
Depositing User:	Esin Yavuz
Date Deposited:	20 Jun 2016 13:54
Last Modified:	24 Oct 2016 09:16
URI:	http://srodev.sussex.ac.uk/id/eprint/61572

View download statistics for this item

📧 Request an update