WorldCat Identities

Kerns, Kelly Michael 1968-

Overview
Works: 1 works in 3 publications in 1 language and 4 library holdings
Genres: Academic theses 
Roles: Author
Publication Timeline
.
Most widely held works by Kelly Michael Kerns
A self-organized data mining agent framework to dynamically discover neural networks by Kelly Michael Kerns( )

3 editions published between 2004 and 2005 in English and held by 4 WorldCat member libraries worldwide

The discipline of artificial intelligence engages machines to emulate human thought. A sub-discipline of artificial intelligence called machine learning involves building memory-reinforced models from training data sets. One approach utilizes a mechanism called a neural network. These networks are ideally suited for solving problems where many examples of outcomes are available, yet it is difficult to formalize a rules-based system for result prediction. This work proposes a tool for generating neural networks from an automated data mining process. The tool itself is a learning machine, built using massive parallelism, self-organization and reinforced memory models. Independent specialized agents cooperate to discover attribute correlations based on Bayesian statistical inference and fundamental database properties. Candidate attributes are extracted into training sets upon which neural networks are dynamically generated and tested. The tool was successfully implemented and tested against different databases, resulting in the production of several useful neural networks
 
Audience Level
0
Audience Level
1
  Kids General Special  
Audience level: 0.56 (from 0.56 for A self-org ... to 0.56 for A self-org ...)

WorldCat IdentitiesRelated Identities
Languages