Date of Award
6-2010
Document Type
Union College Only
Degree Name
Bachelor of Science
Department
Electrical Engineering
First Advisor
Shane Cotter
Language
English
Keywords
data, apple, information, future, historical
Abstract
Assume there exists information in past data that relates to future trends. This is the case with our physical world. It allows physics models to predict exactly how fast an apple will fall. With historical data in the physical world, such as the apple has been 8m above the surface for a long time and that the apple's stem breaks at t = 0, it is possible to predict where the apple will be in the future, at t > 0. If this assumption is made for nancial data, the resulting trend data would be extremely valuable. However there is a great volume of past data and much more complex and transient rules; in short decoding the limited information is di cult. A software signal processing system is proposed to attempt to decode trend information from historical nancial data.
Recommended Citation
Brown, David D., "Evolutionary signal processing : financial time series predictions" (2010). Honors Theses. 1103.
https://digitalworks.union.edu/theses/1103