Date of Award

8-11-2020

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Neuroscience Institute

First Advisor

Yaroslav Molkov

Second Advisor

Igor Belykh

Third Advisor

Vladimir Bondarenko

Fourth Advisor

Andrey Shilnikov

Abstract

In this dissertation, a neurofunctional theory of learning is presented as an extension of functional analysis. This new theory clarifies the distinction— via applied quantitative analysis— between functionally intrinsic (essential) mechanistic structures and irrelevant structural details. This thesis is supported by a review of the relevant literature to provide historical context and sufficient scientific background. Further, the scope of this thesis is elucidated by two questions that are posed from a neurofunctional perspective— (1) how can specialized neuromorphology contribute to the functional dynamics of neural learning processes? (2) Can large-scale neurofunctional pathways emerge via inter-network communication between disparate neural circuits? These questions motivate the specific aims of this dissertation. Each aim is addressed by posing a relevant hypothesis, which is then tested via a neurocomputational experiment. In each experiment, computational techniques are leveraged to elucidate specific mechanisms that underlie neurofunctional learning processes. For instance, the role of specialized neuromorphology is investigated via the development of a computational model that replicates the neurophysiological mechanisms that underlie cholinergic interneurons’ regulation of dopamine in the striatum during reinforcement learning. Another research direction focuses on the emergence of large-scale neurofunctional pathways that connect the cerebellum and basal ganglia— this study also involves the construction of a neurocomputational model. The results of each study illustrate the capability of neurocomputational models to replicate functional learning dynamics of human subjects during a variety of motor adaptation tasks. Finally, the significance— and some potential applications— of neurofunctional theory are discussed.

DOI

https://doi.org/10.57709/18699334

File Upload Confirmation

1

Share

COinS