An investigation of observed Algorithmic Specified Complexity.

Two experiments were carried out to investigate how Algorithmic Specified Complexity (ASC) might serve as a tool, specifically in the area of classification AI, and how well the theory around it predicts the characteristics of random numbers. One evaluated an approach to measuring ASC in pictures by how well it helped in classification, and the other compared predictions and observations of the compressibility of random bitstrings. The ASC of MNIST pictures was estimated by saving concatenations of samples as PNG. The expected ASC of random bitstrings was compared to average observed ASC (OASC) values from LZ78 Huffman codes. Observed ASC of MNIST pictures helped to identify them, and as predicted, expectations of ASC were higher than those of OASC. ASC shows value in AI applications, and while generic compression algorithms show some promise, the best way to measure ASC is by functionality.

Name thesis:degree_name

M.S.E.C.E.

Level thesis:degree_level

Masters

Grantor

Baylor University.

Year dc:date.issued

2017

Author dc:creator

• Nemati, David C., 1989-

Advisor dc:contributor.advisor

• Marks, Robert J., II (Robert Jackson), 1950-

• Information theory.
• Algorithmic Specified Complexity.

Statement dc:rights

• Baylor University works are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. Contact libraryquestions@baylor.edu for inquiries about permission.

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/2104/10179

OAI identifier oai:identifier

oai:baylor-ir.tdl.org:2104/10179