지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
논문 기본 정보
- 자료유형
- 학술저널
- 저자정보
- 발행연도
- 2013.8
- 수록면
- 103 - 115 (13page)
이용수
초록· 키워드
오류제보하기
A comparative study has been done in attempt to evaluate some of the major properties of weighted mixtures of distributions and their predictive ability for risk management purposes. The process of constructing weighted sums of normal distributions has been demonstrated starting from the properties of the most common distributions, continuing with problems that lead to combining individual distributions in mixtures, and finishing with analysis of the predictive ability of two different brands of mixtures: arithmetic and weighted sum of normal distributions. The models was applied to three sets of economic data for South Korea. The estimation of predictive ability in these cases was based on their discrimination ability. The results obtained after performing linear discriminant analysis show better accuracy (between 0.28% and 9.5%) of the weighted sums of normal distributions than the one of the arithmetic sums.
목차
Abstract
I. Introduction
II. Some Statistical Properties of the Most Common Distributions
III. Mixtures of Statistical Distributions
IV. Experiments and Results
V. Discussion and Conclusion
References
참고문헌 (0)
참고문헌 신청이 논문의 저자 정보
최근 본 자료
댓글(0)
0