Showing 1–2 of 2 results for author: Galtier, M

A biological gradient descent for prediction through a combination of STDP and homeostatic plasticity

Authors: Mathieu Galtier, Gilles Wainrib

Abstract: Identifying, formalizing and combining biological mechanisms which implement known brain functions, such as prediction, is a main aspect of current research in theoretical neuroscience. In this letter, the mechanisms of Spike Timing Dependent Plasticity (STDP) and homeostatic plasticity, combined in an original mathematical formalism, are shown to shape recurrent neural networks into predictors. F… ▽ More Identifying, formalizing and combining biological mechanisms which implement known brain functions, such as prediction, is a main aspect of current research in theoretical neuroscience. In this letter, the mechanisms of Spike Timing Dependent Plasticity (STDP) and homeostatic plasticity, combined in an original mathematical formalism, are shown to shape recurrent neural networks into predictors. Following a rigorous mathematical treatment, we prove that they implement the online gradient descent of a distance between the network activity and its stimuli. The convergence to an equilibrium, where the network can spontaneously reproduce or predict its stimuli, does not suffer from bifurcation issues usually encountered in learning in recurrent neural networks. △ Less

Submitted 11 June, 2013; v1 submitted 21 June, 2012; originally announced June 2012.

Comments: 36 pages, 3 figures

On an explicit representation of the solution of linear stochastic partial differential equations with delays

Authors: Mathieu Galtier, Jonathan Touboul

Abstract: Based on the analysis of a certain class of linear operators on a Banach space, we provide a closed form expression for the solutions of certain linear partial differential equations with non-autonomous input, time delays and stochastic terms, which takes the form of an infinite series expansion. Based on the analysis of a certain class of linear operators on a Banach space, we provide a closed form expression for the solutions of certain linear partial differential equations with non-autonomous input, time delays and stochastic terms, which takes the form of an infinite series expansion. △ Less

Submitted 7 September, 2011; v1 submitted 30 March, 2011; originally announced March 2011.