| Author | Yang, Jung-hsiung |
| Call Number | AIT Thesis no. CS-87-15 |
| Subject(s) | Flood control--Computer programs
|
| Note | A thesis submitter 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 Log Pearson Type 3 (LP3) distribution has been widely
used in flood frequency analysis, which is based upon complete
samples of annual flood data. Due to the inaccuracy in flood data
measurement, many data sets contain outliers or unreliable
values. This situation calls for the application of censoring
theory to provide appropriate treatments.
In this study, the maximum likelihood method CMML> was used
to estimate the LP3 parameters and hence quantiles, represented
by the T-year events, from type I censored samples. All needed
expressions for the ML equations were derived and formulas for
computing the asymptotic Variance-covariance matrix of the
parameter estimators were established. Related computer programs
were subsequently developed which were the used in extensive
Monte Carlo experiments to determine the effects of censoring on
the ML estimators. It was found that for the sets of parameter
values considered, the left censoring produced estimators with
substantial reduction in bias for positive scale values.
Moreover, it was also found that right censoring can be used to
deal with the problem of large outliers. In this case, right
censoring is more advantageous than discarding the outlier which
is in turn better the taking the outlier into account. |
| Year | 1987 |
| 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) | Khang, Bach Hung ;Venkatachalam, Paruvachi A.
|
| Scholarship Donor(s) | The Government of Republic of China |
| Degree | Thesis (M.Sc.) - Asian Institute of Technology, 1987 |