You are here

CHARACTERIZATION OF AN ADVANCED NEURON MODEL

Download pdf | Full Screen View

Date Issued:
2012
Abstract/Description:
This thesis focuses on an adaptive quadratic spiking model of a motoneuron that is both versatile in its ability to represent a range of experimentally observed neuronal firing patterns as well as computationally efficient for large network simulation. The objective of research is to fit membrane voltage data to the model using a parameter estimation approach involving simulated annealing. By manipulating the system dynamics of the model, a realizable model with linear parameterization (LP) can be obtained to simplify the estimation process. With a persistently excited current input applied to the model, simulated annealing is used to efficiently determine the best model parameters that minimize the square error function between the membrane voltage reference data and data generated by the LP model. Results obtained through simulation of this approach show feasibility to predict a range of different neuron firing patterns.
Title: CHARACTERIZATION OF AN ADVANCED NEURON MODEL.
24 views
14 downloads
Name(s): Echanique, Christopher, Author
Behal, Aman, Committee Chair
University of Central Florida, Degree Grantor
Type of Resource: text
Date Issued: 2012
Publisher: University of Central Florida
Language(s): English
Abstract/Description: This thesis focuses on an adaptive quadratic spiking model of a motoneuron that is both versatile in its ability to represent a range of experimentally observed neuronal firing patterns as well as computationally efficient for large network simulation. The objective of research is to fit membrane voltage data to the model using a parameter estimation approach involving simulated annealing. By manipulating the system dynamics of the model, a realizable model with linear parameterization (LP) can be obtained to simplify the estimation process. With a persistently excited current input applied to the model, simulated annealing is used to efficiently determine the best model parameters that minimize the square error function between the membrane voltage reference data and data generated by the LP model. Results obtained through simulation of this approach show feasibility to predict a range of different neuron firing patterns.
Identifier: CFH0004259 (IID), ucf:44958 (fedora)
Note(s): 2012-08-01
B.S.P.E.
Engineering and Computer Science, Dept. of Electrical Engineering and Computer Science
Bachelors
This record was generated from author submitted information.
Subject(s): neuron
model
parameter estimation
Izhikevich
simulated annealing
spiking
Persistent Link to This Record: http://purl.flvc.org/ucf/fd/CFH0004259
Restrictions on Access: campus 2013-08-01
Host Institution: UCF

In Collections