AComparativeStudyonLinearRegressionMethodsaboutBinomialParameter
碩士 === 國立中正大學 === 數理統計研究所 === 93 === This paper describes a computational procedure for obtaining maximum likelihood estimators (MLE) in linear models regarding binary parameter. The maximum likelihood estimators (MLE) are compared with the corresponding least square estimators (LSE) in regard to th...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2005
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/37gfb3 |
| Summary: | 碩士 === 國立中正大學 === 數理統計研究所 === 93 === This paper describes a computational procedure for obtaining maximum likelihood estimators (MLE) in linear models regarding binary parameter. The maximum likelihood estimators (MLE) are compared with the corresponding least square estimators (LSE) in regard to their performance in accuracy of estimation. The comparison study covers normal regression, binomial regression, and Poisson regression. Computer simulations are properly undertaken to accomplish the comparison effort regarding MLE and LSE. The focused concern is about cases when the binary parameter is small while the sample size is very large.
|
|---|