|
Nelineinaya Dinamika [Russian Journal of Nonlinear Dynamics], 2011, Volume 7, Number 4, Pages 859–875
(Mi nd297)
|
|
|
|
This article is cited in 2 scientific papers (total in 2 papers)
Reinforcement learning of a spiking neural network in the task of control of an agent in a virtual discrete environment
O. Yu. Sinyavskiy, A. I. Kobrin National Research University Moscow Power Engineering Institute, Krasnokazarmennaya st. 14, Moscow, 111250, Russia
Abstract:
Method of reinforcement learning of spiking neural network that controls robot or virtual agent is described. Using spiking neurons as key elements of a network allows us to exploit spatial and temporal structure of input sensory information. Teaching of the network is implemented with a use of reinforcement signals that come from the external environment and reflect the success of agents recent actions. A maximization of the received reinforcement is done via modulated minimization of neurons informational entropy that depends on neurons weights. The laws of weights changes were close to modulated synaptic plasticity that was observed in real neurons. Reinforcement learning algorithm was tested on a task of a resource search in a virtual discrete
environment.
Keywords:
spiking neuron, adaptive control, reinforcement learning, informational entropy.
Received: 06.06.2011 Accepted: 30.09.2011
Citation:
O. Yu. Sinyavskiy, A. I. Kobrin, “Reinforcement learning of a spiking neural network in the task of control of an agent in a virtual discrete environment”, Nelin. Dinam., 7:4 (2011), 859–875
Linking options:
https://www.mathnet.ru/eng/nd297 https://www.mathnet.ru/eng/nd/v7/i4/p859
|
Statistics & downloads: |
Abstract page: | 1037 | Full-text PDF : | 928 | References: | 74 | First page: | 1 |
|