Content area
Abstract
This thesis explores the use of a relatively new field in the artificial intelligence area called Rough Set Theory to the application of economic and stock market data. The theory of Rough Sets is reviewed along with alternative AI methods. A learning experiment is set up using representative condition attributes from the computer, semiconductor, and semiconductor equipment industries. The program DataLogic/R+ is used as the rule generator. Computer programs are written to preprocess and discretize the time series data. The effects of averaging, delaying, and summing are explored. The propositional rules generated are then validated using recent data. The results show that strong predictive rules can be generated.
Details
Title
An application of rough sets to economic and stock market data
Author
Tremba, Joseph Alexander
Year
1997
ISBN
978-0-591-70732-8
Source type
Dissertation or Thesis
Language of publication
English
ProQuest document ID
304414835
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
