Forecasting of seasonal streamflows | |
| Author | Chen, Ming-an |
| Call Number | AIT Thesis no. CS-94-10 |
| Subject(s) | Regression analysis Hydraulics--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 | In this study, the time series of monthly flows of the Mekong River and ten-day flows of the Tachia River were used to build forecasting models by using multiple regression analysis, BoxJenkins method and back propagation method. All of the forecasting methods with lead time = 1 and 2 were implemented in this study. It was found that back propagation method can obtain the best results during calibration period and Box-Jenkins models can get better results than multiple regression method in most of the cases. For forecasting seasonal flows by back propagation method, using one hidden layer network is enough to get quite satisfactory results. The forecasting accuracy for lead time= 2 (time units) is a little worse than that for lead time= 1 as expected. |
| Year | 1994 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Other Field of Studies (No Department) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Huynh Ngoc Phien |
| Examination Committee(s) | Do Ba Khang ;Gonzales, Jr. , R. L. |
| Scholarship Donor(s) | R.O.C. Government |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1994 |