Applied Math Seminar | Marius Yamakou, A simple parameter can switch between different weak-noise–induced phenomena in neurons

Thursday, February 16, 2023 2:30 pm - 2:30 pm EST (GMT -05:00)
Export this event to calendar
iCal

MC 6460
Ìý

Speaker

Marius Yamakou, Friedrich-Alexander-Universität Erlangen-Nürnberg

Title

A simple parameter can switch between different weak-noise–induced phenomena in neurons

Abstract

This talkÌýwillÌýconsider a stochastic multiple-timescaleÌýdynamical system modeling a biological neuron. WithÌýthis model, we willÌýseparatelyÌýuncover theÌýmechanisms underlyingÌýtwo different ways biological neurons encode information with stochastic perturbations:Ìýself-induced stochastic resonance (SISR) and inverse stochastic resonance (ISR).Ìý ÌýWe will thenÌýshow that in the same weak noise limit,ÌýSISR and ISR areÌýrelated through the relative geometric positioning (and stability) of the fixedÌýpoint and the genericÌýfolded singularity of the model'sÌýcritical manifold.ÌýÌý

This mathematical resultÌýcouldÌýexplainÌýtheÌýexperimentalÌýobservationÌýwhere neurons with identical morphological features sometimesÌýencode different information with identical synaptic input. Finally, if time permits, we shall discuss the plausibleÌýapplications of this result in neuro-biologically inspired machine learning, particularly reservoir computing based on liquid-state machines.