A Study of AIC, SICc, and Cp in Non-Orthogonal Regression

• 발행기관: 한국자료분석학회
• 저자명: Chan Keun Park
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.4 No.4, 373~386쪽, 전체 14쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2002.12.30

영문 초록

There are many model selection criteria in regression analysis. We choose AIC, SICc, and Cp and compare the power of these model selection criteria to find the true model when the independent variables are non-orthogonal. We already compared the power of these criteria in orthogonal regression case (Park and Yoda, 2000, Park 2000). We decide the correlation as 0.4.

목차

1. Introduction

2. Survey of literature review

3. Simulation Study

4. Non-orthogonal case

5. Conclusion and Further Research

References

키워드

SSE AIC SICc Cp Non-orthogonal regression