| Author | Cruda, Rodolfo B. |
| Call Number | AIT Thesis no. WA-83-3 |
| Subject(s) | Water resources development--Data processing
|
| Note | A thesis submitted in partial fulfillment of the requirement for the degree of Master of Engineering, school of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | An empirical heuristic self-organization approach, referred to as the group method of data handling (GMDH) , was modified and used to construct nonlinear models to estimate and forecast evapotranspiration in the Red
River Del ta and streamflow in Ping River and Huai Sato River Basins. The method used principal components in place of the original variables as input
variables to the modified algorithm. In t his algorithm, the partial model quality criterion, such as MSE (Mean Square Error), ISE (Integral Square
Error) , or MAD (Mean Absolute Deviation) , was utilized as a criterion for selecting intermediate variables. The modifications resulted in a reduced number of terms in the final model equations, a decrease in multicollinearity
problem, and a lesser computational difficulties. This method appears to be
advantageous with systems characterized by complexity with many variables and parameters, not well-defined mathematical structures, and limited data. In particular, t h is algorithm is useful in generating hypotheses about the important variables and structures of the system. |
| Year | 1982 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Civil and Infrastucture Engineering (DCIE) |
| Academic Program/FoS | Water Resources Research Engineering (WA) |
| Chairperson(s) | Huynh Ngoc Phien |
| Examination Committee(s) | Hoshi, Kiyoshi ; Suphat Vongvisessomjai
|
| Scholarship Donor(s) | Royal Thai Government |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1982 |