Using data mining techniques to identify “the best” operational patterns for enrollment modeling

For any Educational Institution it is very important to know the number of new students and the number of returning students. Based on these numbers, there could be conducted predictions of the budget that the institution will have for the next year. This research will utilize pre-existing historical data from Texas Woman's University containing readily available and easily measured factors, which most institutions of higher learning will have available, and will split the existing data in all the sub sets possible. Running a chi square analysis on each set obtained, the program will be able to show us which splitting way is better for obtaining the most consistent patterns, using the provided data. The results will be compared with the results obtained running a linear regression analysis on the same data sets. The study will introduce an extraneous hidden-time variable related to partitioning ways possible. The program can be used in the future on any University data sets, providing the most holding combination of variables that will hold over the years.

Name thesis:degree_name

Master of Science

Level thesis:degree_level

Master

Discipline thesis:degree_discipline

Mathematics

Grantor

Texas Woman's University

Year dc:date.issued

2009

Author dc:creator

• Obarse, Bogdan Catalin

Chair dc:contributor.committeechair

• Hamner, Mark S.

Committee members dc:contributor.committeemember

• Marshall, David, Ph. D.
• Grigorieva, Ellina

• Education
• Applied sciences
• Pure sciences

Language dc:language.iso

en_US

Handle dc:identifier.uri

https://hdl.handle.net/11274/10586

OAI identifier oai:identifier

oai:twu-ir.tdl.org:11274/10586