| Author | Ng, Kwang Iak |
| Call Number | AIT SSPR no.CA-82-10 |
| Subject(s) | Regression analysis--Computer programs
|
| Note | A special study submitted in partial fulfillment of the requirements for the degree of Master of Science, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | The study deals with stepwise procedure of selecting subset of regressor variables and a brief review of two popular methods to overcome undesirable behaviors of least squares estimation of regression coefficients when regressor variables exhibit mild or strong multicollinearity.
In stepwise regression , the use of AIC as selecting criterion is investigated. It is found that AIC does not differ from other criteria. The serious effects of
arbitrary chosen F Values on the ways the selection procedure behaves arc illustrated . A solution for automatic generation of F values based on reduction factor SSE(p)/SSE(p-1) is also presented.
An attempt has been made to compare the performance of ridge regression and regression on principal components to overcome multicollinearity. The comparison statistics are obtained from sample data simulated by Owen and Steck scheme.
The developed computer programs for this study include
stepwise regression program (STPRGA), ridge regression
program (RIDGE) for respective k values, program to simulate
sample data (SIMU), predictive performance and reporting
programs for obtaining the results of analysis. The distinct
feature of STPRGA is its capability to include ridge
regression analysis and has rooms tor many useful features
to be added .
|
| Year | 1982 |
| Type | Special Study Project Report (SSPR) |
| School | School of Engineering and Technology |
| Department | Department of Information and Communications Technologies (DICT) |
| Chairperson(s) | Huynh, Ngoc Phien |
| Examination Committee(s) | Malik, Ghulam M. ; Vilas Wuwongse
|
| Scholarship Donor(s) | The Australian Government |
| Degree | Special Studies Project Report (M. Sc.) - Asian Institute of Technology, 1982 |