로오버스트 位置母數 推定量에 관한 擬態分析硏究

• 영문명: A Simulation Study on the Robust Estimates of Location
• 발행기관: 건국대학교 경제경영연구소
• 저자명: Sang Ik Kim(金商益)
• 간행물 정보: 『상경연구』제15권, 127~141쪽, 전체 15쪽
• 주제분류: 경제경영 > 경제학
• 파일형태: PDF
• 발행일자: 1990.09.15

영문 초록

The robust estimation procedures for the location parameter have been studied extensively in recent years. These estimates are useful when the underlying distribution of sample data is not exactly known or the sample data contains some outliers. In this paper we compare the different robust estimates of location parameter by simulation analysis. In stead of using large amount of data as other previous papers, we use the smaller size data to compare. The results show that the appropriately trimmed mean and Jaeckel’s adaptive estimate are the best with regard to bias and variance of the estimate. However, the sample median is not as good as others since it has larger variance even though it has small bias.

목차

Ⅰ. 序論

Ⅱ. 로오버스트 推定方法

Ⅲ. 擬態分析

Ⅳ. 結論

參考文獻

SUMMARY