異分散的 誤差項을 갖는 非線型回歸模型에 있어서 誤差項表記에 대한 檢定

• 영문명: Testing for the error specification in nonlinear regression models with hereroskedastic errors
• 발행기관: 건국대학교 경제경영연구소
• 저자명: Seok Koo Lee(李錫求)
• 간행물 정보: 『상경연구』제28권 제1호, 1~11쪽, 전체 11쪽
• 주제분류: 경제경영 > 경제학
• 파일형태: PDF
• 발행일자: 2003.05.30

영문 초록

This paper deals with the problem of appropriately specifying the error structure when error terms are heteroskedastic in nonlinear regression models. The Box-Cox transformation model is presented which includes disparate hypotheses of additive normal and multiplicative lognormal error distributions. Also, the likelihood-ratio test for selecting the appropriate error specification is suggested. And a simple example is given in which the developed method is applied in the context of estimating a nonlinear Engel curve.

목차

Ⅰ. 序論

Ⅱ. 模型設定

Ⅲ. 尤度比檢定

Ⅳ. 하나의 例

Ⅴ. 結論

참고문헌

Abstract