Dissertation/Thesis Abstract

A field-based artificial neural network with cerebellar model for complex motor sequence learning
by Brady, Michael Connolly, Ph.D., Indiana University, 2012, 187; 3509892
Abstract (Summary)

This dissertation addresses two related questions: (1) what form might a complex motor plan take in the brain, and (2) how might such a plan be converted into coordinated motor behavior? One approach to modeling sophisticated human motor skills, like typing or musical performance or speech is to conceptualize the motor plan as a stream of discrete instructions. This approach has perhaps been attractive in cognitive science because it allows theorists to discuss motor plans in terms of symbolic notations. However, many brain and robotics researchers have come to view the concept of the symbolic command sequence as motor plan to be wanting. This dissertation offers a concrete alternative based on established brain circuitry. It demonstrates and analyzes how a control signal for complex motor behavior through time may be modeled as a composite of neural firing patterns. Activation patterns that correspond to a production sequence arrive from multiple sources and persist in tandem over relatively long spans of time as they act to influence and contextualize each other. Such persistent and holistic activity must then be transformed into serial behavior. This is achieved through neural dynamics. Motor activity emerges as a result of interactions between long-time resonance patterns and short-time worldly input and feedback. Modeling involves the cortico-thalamic circuit together with a modern depiction of the functional role of the cerebellum. The cerebellar model provides a temporally structured signal that helps to inform cortico-thalamic processing and motor rehearsal. The system is situated in a speech motor control and feedback framework. The control framework includes a mechanical vocal tract and the dissertation as a whole may be interpreted as the multidisciplinary foundation for a speech robotics research program.

Indexing (document details)
Advisor: Beer, Randall, Kewley-Port, Diane
Commitee: Bingham, Geoffrey, Port, Robert
School: Indiana University
Department: Cognitive Science
School Location: United States -- Indiana
Source: DAI-B 73/10(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Neurosciences, Robotics
Keywords: Artificial neural networks, Complex motors, Motor plans, Sequence learning
Publication Number: 3509892
ISBN: 978-1-267-36964-2
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest