A neural network for coding of trajectories by time series of neuronal population vectors

Authors:
Alexander V. Lukashin;Apostolos P. Georgopoulos
Affiliations:
-;-
Venue:
Neural Computation
Year:
1994

Citing 0
Cited 5

Self-organization of firing activities in monkey's motor cortex: trajectory computation from spike signals

Neural Computation
Linking perception and neurophysiology for motion pattern processing: the computational power of inhibitory connections in cortex

Optic flow and beyond
Dynamical Mechanism for Sharp Orientation Tuning in an Integrate-and-Fire Model of a Cortical Hypercolumn

Neural Computation
A Model of Reaching Dynamics in Primary Motor Cortex

Journal of Cognitive Neuroscience
Learning Complex Population-Coded Sequences

ICANN '09 Proceedings of the 19th International Conference on Artificial Neural Networks: Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

The neuronal population vector is a measure of the combineddirectional tendency of the ensemble of directionally tuned cellsin the motor cortex. It has been found experimentally that atrajectory of limb movement can be predicted by adding togetherpopulation vectors, tip-to-tail, calculated for successive instantsof time to construct a neural trajectory. In the present paper weconsider a model of the dynamic evolution of the population vector.The simulated annealing algorithm was used to adjust the connectionstrengths of a feedback neural network so that it would generate agiven trajectory by a sequence of population vectors. This wasrepeated for different trajectories. Resulting sets of connectionstrengths reveal a common feature regardless of the type oftrajectories generated by the network: namely, the mean connectionstrength was negatively correlated with the angle between thepreferred directions of neuronal pair involved in the connection.The results are discussed in the light of recent experimentalfindings concerning neuronal connectivity within the motorcortex.