University of Notre Dame
Browse
- No file added yet -

Dynamics and Computations in Recurrent Neural Networks

Download (7.77 MB)
thesis
posted on 2019-05-14, 00:00 authored by Ryan Pyle

Randomly connected recurrent neural networks (RNNs) serve as a parsimonious model of cortical dynamics, and could be used to model memory, decision making, and cognition. Machine-learning based variants of RNNs have recently gained popularity due to their utility in a wide array of application including speech recognition, medical outcome prediction, handwriting recognition, or robot control. However, the methods used in these applications are either too far removed from biological RNNs, or not biologically plausible. A biologically plausible method could both provide insight into how biological RNNs function, and improve performance in artificial systems.

Here I describe my contributions towards biologically plausible algorithms for RNNs. I begin with an extension of balanced network theory, which creates a parsimonious model of neural dynamics. Existing RNN algorithms use simplified neuron models due to difficulties using more complex or realistic neuron models. Accounting for the spatially dependent structure observed in real cortical networks increases the reliability of the reservoir network, allowing it work in realistic spiking networks. Finally, I develop a biologically-inspired RNN algorithm, which solves an issue of unrealistic supervision found in most existing algorithms.

History

Date Modified

2019-08-08

Defense Date

2019-04-16

CIP Code

  • 27.9999

Research Director(s)

Robert J. Rosenbaum

Committee Members

Alan Lindsay Lizhen Lin

Degree

  • Doctor of Philosophy

Degree Level

  • Doctoral Dissertation

Alternate Identifier

1111684901

Library Record

5172328

OCLC Number

1111684901

Program Name

  • Applied and Computational Mathematics and Statistics

Usage metrics

    Dissertations

    Categories

    No categories selected

    Keywords

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC