An investigation in size reduction in neural networks

Graduate Thesis Or Dissertation

Public Deposited

Citeable URL: https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/ns064835h

Descriptions

Attribute Name	Values
Creator	Warren, Andrew Howard
Abstract	In this thesis, the reduction of neural networks is studied. A new, largely automatic method is developed for reducing a neural network, and its capabilities are analyzed and compared with current methods. An example is presented that is irreducible via existing techniques, but is reducible by the new method. Factors affecting convergence to solutions using the back propagation learning method are discussed. In addition, the heuristic of adding hidden units to a neural network to obtain convergence is analyzed mathematically.
Resource Type	Masters Thesis
Date Available	2013-07-22T16:29:10+00:00
Date Issued	1989-08-15
Degree Level	Master's
Degree Name	Master of Science (M.S.)
Degree Field	Mathematics
Degree Grantor	Oregon State University
Commencement Year	1990
Advisor	Burton, Robert M.
Academic Affiliation	Mathematics
Non-Academic Affiliation	Oregon State University. Graduate School
Subject	Heuristic programming Artificial intelligence Neural circuitry
Rights Statement	Copyright Not Evaluated
Publisher	Oregon State University
Peer Reviewed	No
Language	English [eng]
Digitization Specifications	File scanned at 300 ppi (Monochrome) using ScandAll PRO 1.8.1 on a Fi-6770A in PDF format. CVista PdfCompressor 5.0 was used for pdf compression and textual OCR.
Replaces	http://hdl.handle.net/1957/40921

Parents:

This work has no parents.

In Collection:

Thumbnail	Title	Date Uploaded	Visibility	Actions
	WarrenAndrewHoward1990.pdf	2017-08-04	Public	Download