• 제목/요약/키워드: Hellinger distance

검색결과 23건 처리시간 0.02초

Negative Exponential Disparity Based Deviance and Goodness-of-fit Tests for Continuous Models: Distributions, Efficiency and Robustness

  • Jeong, Dong-Bin;Sahadeb Sarkar
    • Journal of the Korean Statistical Society
    • /
    • 제30권1호
    • /
    • pp.41-61
    • /
    • 2001
  • The minimum negative exponential disparity estimator(MNEDE), introduced by Lindsay(1994), is an excellenet competitor to the minimum Hellinger distance estimator(Beran 1977) as a robust and yet efficient alternative to the maximum likelihood estimator in parametric models. In this paper we define the negative exponential deviance test(NEDT) as an analog of the likelihood ratio test(LRT), and show that the NEDT is asymptotically equivalent to he LRT at the model and under a sequence of contiguous alternatives. We establish that the asymptotic strong breakdown point for a class of minimum disparity estimators, containing the MNEDE, is at least 1/2 in continuous models. This result leads us to anticipate robustness of the NEDT under data contamination, and we demonstrate it empirically. In fact, in the simulation settings considered here the empirical level of the NEDT show more stability than the Hellinger deviance test(Simpson 1989). The NEDT is illustrated through an example data set. We also define a goodness-of-fit statistic to assess adequacy of a specified parametric model, and establish its asymptotic normality under the null hypothesis.

  • PDF

Hellinger 거리 IoU와 Objectron 적용을 기반으로 하는 객체 감지 (Object Detection Based on Hellinger Distance IoU and Objectron Application)

  • 김용길;문경일
    • 한국인터넷방송통신학회논문지
    • /
    • 제22권2호
    • /
    • pp.63-70
    • /
    • 2022
  • 2D 객체 감지 시스템은 최근 몇 년 동안 심층 신경망과 대규모 이미지 데이터세트의 사용으로 크게 개선되었지만, 아직도 범주 내에서 데이터 부족, 다양한 외관 및 객체 형상 때문에 자율 탐색 등과 같은 로봇 공학과 관련된 응용에서 2D 물체 감지 시스템은 적절하지 않다. 최근에 소개되고 있는 구글 Objectron 또한 증강 현실 세션 데이터를 사용하는 새로운 데이터 파이프라인이라는 점에서 도약이라 할 수 있지만, 3D 공간에서 2D 객체 이해라는 측면에서 마찬가지로 한계가 있다. 이에 본 연구에서는 더 성숙한 2D 물체 감지 방법을 Objectron에 도입하는 3D 물체 감지 시스템을 나타낸다. 대부분의 객체 감지 방법은 경계 상자를 사용하여 객체 모양과 위치를 인코딩한다. 본 작업에서는 가우스 분포를 사용하여 객체 영역의 확률적 표현을 탐색하는데, 일종의 확률적 IoU라 할 수 있는 Hellinger 거리를 기반으로 하는 가우스 분포에 대한 유사성 측도를 제시한다. 이러한 2D 표현은 모든 객체 감지기에 원활하게 통합할 수 있으며, 실험 결과 데이터 집합에서 주석이 달린 분할 영역에 더 가까워서 Objectron의 단점이라 할 수 있는 3D 감지 정확도를 높일 수 있다.

Empirical Comparisons of Disparity Measures for Three Dimensional Log-Linear Models

  • Park, Y.S.;Hong, C.S.;Jeong, D.B.
    • Journal of the Korean Data and Information Science Society
    • /
    • 제17권2호
    • /
    • pp.543-557
    • /
    • 2006
  • This paper is concerned with the applicability of the chi-square approximation to the six disparity statistics: the Pearson chi-square, the generalized likelihood ratio, the power divergence, the blended weight chi-square, the blended weight Hellinger distance, and the negative exponential disparity statistic. Three dimensional contingency tables of small and moderate sample sizes are generated to be fitted to all possible hierarchical log-linear models: the completely independent model, the conditionally independent model, the partial association models, and the model with one variable independent of the other two. For models with direct solutions of expected cell counts, point estimates and confidence intervals of the 90 and 95 percentage points of six statistics are explored. For model without direct solutions, the empirical significant levels and the empirical powers of six statistics to test the significance of the three factor interaction are computed and compared.

  • PDF

Minimum Disparity Estimation for Normal Models: Small Sample Efficiency

  • Cho M. J.;Hong C. S.;Jeong D. B.
    • Communications for Statistical Applications and Methods
    • /
    • 제12권1호
    • /
    • pp.149-167
    • /
    • 2005
  • The minimum disparity estimators introduced by Lindsay and Basu (1994) are studied empirically. An extensive simulation in this paper provides a location estimate of the small sample and supplies empirical evidence of the estimator performance for the univariate contaminated normal model. Empirical results show that the minimum generalized negative exponential disparity estimator (MGNEDE) obtains high efficiency for small sample sizes and dominates the maximum likelihood estimator (MLE) and the minimum blended weight Hellinger distance estimator (MBWHDE) with respect to efficiency at the contaminated model.

A Note on Smoothing Distribution Function Estimation

  • Chu, In-Sun;Choi, Jae-Ryong
    • Communications for Statistical Applications and Methods
    • /
    • 제4권3호
    • /
    • pp.911-915
    • /
    • 1997
  • The purpose of this paper is to consider the problem of selection of optimal smoothing parameter for kernel-type distribution function estimator, which asymptotically minimizes mean Hellinger distance.

  • PDF

Tests of Hypotheses in Multiple Samples based on Penalized Disparities

  • Park, Chanseok;Ayanendranath Basu;Ian R. Harris
    • Journal of the Korean Statistical Society
    • /
    • 제30권3호
    • /
    • pp.347-366
    • /
    • 2001
  • Robust analogues of the likelihood ratio test are considered for testing of hypotheses involving multiple discrete distributions. The test statistics are generalizations of the Hellinger deviance test of Simpson(1989) and disparity tests of Lindsay(1994), obtained by looking at a 'penalized' version of the distances; harris and Basu (1994) suggest that the penalty be based on reweighting the empty cells. The results show that often the tests based on the ordinary and penalized distances enjoy better robustness properties than the likelihood ratio test. Also, the tests based on the penalized distances are improvements over those based on the ordinary distances in that they are much closer to the likelihood ratio tests at the null and their convergence to the x$^2$ distribution appears to be dramatically faster; extensive simulation results show that the improvement in performance of the tests due to the penalty is often substantial in small samples.

  • PDF

베타다양성 개념의 적용을 통한 청계천 어류 군집 특성 분석 (Application of Beta Diversity to Analysis the Fish Community Structure in Stream)

  • 김동환;이완옥;홍양기;전형주;김경환;강혜진;송미영
    • 생태와환경
    • /
    • 제52권3호
    • /
    • pp.274-283
    • /
    • 2019
  • 청계천에 서식하는 어류 군집의 공간적 변이와 환경과의 관계를 측정하기 위해, 청계천 내 6개 지점을 대상으로 2년간(2014~2015년) 이화학적 요인, 서식처 환경, 어류 군집을 조사하였다. 어류 군집의 공간적 변이는 지점-종 군집 데이터 메트릭스를 기반으로 한 베타다양성 분석을 통해 정량적으로 제시하였다. 또한 청계천 내 전체 군집 변이 값(베타다양성)과 함께 각각의 지점이 청계천 전체 베타다양도에 기여하는 값 (LCBD, Local Contribution to Beta Diversity)도 계산하였다. 데이터 분석의 기반이 되는 지점-종 군집 데이터 테이블은 출현-비출현, 풍부도, 헤링거 변환 값의 세 가지 형태로 적용하였고, 해당 데이터 형태에 따른 베타다양성과 지점별 변이 영향을 각각 계산하여 비교하였다. 헤링거 변환을 통해 계산된 베타다양성 값은 출현-비출현 정보나 풍부도를 바탕으로 한 분석보다 큰 값을 보여주어 공간적 변이를 가장 잘 보여주는 것으로 나타났다. 각 지점별 군집 변이 기여도(LCBD)는 출현-비출현 정보와 해링거 변환을 통한 분석이 유사한 경향을 보여주었다. 자료의 정규성을 가지기 어려운 어류 군집 자료의 경우 풍부도를 이용한 공간 변이 분석은 적절하지 않은 것으로 판단된다. 추가적으로 다양한 환경 요인 및 군집 지수와 베타다양성 기여도 값의 관계를 상관분석을 통해 나타내었다. 해당 지점의 알파다양성 지수와 베타다양성 기여도가 높은 음의 상관관계를 보였고 이는 선행 연구와 유사한 결과이다. 본 연구에 적용한 방법은 매트릭스 형태의 자료를 대상으로 베타다양성 계산과 지점별 군집 변이 기여도를 수치화하는 데 유용한 것으로 나타났다.

Bayesian Model Selection in the Unbalanced Random Effect Model

  • Kim, Dal-Ho;Kang, Sang-Gil;Lee, Woo-Dong
    • Journal of the Korean Data and Information Science Society
    • /
    • 제15권4호
    • /
    • pp.743-752
    • /
    • 2004
  • In this paper, we develop the Bayesian model selection procedure using the reference prior for comparing two nested model such as the independent and intraclass models using the distance or divergence between the two as the basis of comparison. A suitable criterion for this is the power divergence measure as introduced by Cressie and Read(1984). Such a measure includes the Kullback -Liebler divergence measures and the Hellinger divergence measure as special cases. For this problem, the power divergence measure turns out to be a function solely of $\rho$, the intraclass correlation coefficient. Also, this function is convex, and the minimum is attained at $\rho=0$. We use reference prior for $\rho$. Due to the duality between hypothesis tests and set estimation, the hypothesis testing problem can also be solved by solving a corresponding set estimation problem. The present paper develops Bayesian method based on the Kullback-Liebler and Hellinger divergence measures, rejecting $H_0:\rho=0$ when the specified divergence measure exceeds some number d. This number d is so chosen that the resulting credible interval for the divergence measure has specified coverage probability $1-{\alpha}$. The length of such an interval is compared with the equal two-tailed credible interval and the HPD credible interval for $\rho$ with the same coverage probability which can also be inverted into acceptance regions of $H_0:\rho=0$. Example is considered where the HPD interval based on the one-at- a-time reference prior turns out to be the shortest credible interval having the same coverage probability.

  • PDF

Robust Discriminant Analysis using Minimum Disparity Estimators

  • 조미정;홍종선;정동빈
    • 한국통계학회:학술대회논문집
    • /
    • 한국통계학회 2004년도 학술발표논문집
    • /
    • pp.135-140
    • /
    • 2004
  • Lindsay and Basu (1994)에 의해 소개된 최소차이추정량 (Minimum Disparity Estimators)들은 실제 자료 분석 도구로써 유용하다. 본 논문에서는 최소일반화음지수 차이추정량 (Minimum Generalized Negative Exponential Disparity Estimator, MGNEDE)이 최대가능도추정량 (Maximum Likelihood Estimator, MLE)와 최소가중 헬링거거리추정량 (Minimum Blended Weight Hellinger Distance Estimator, MBWHDE)에 비해 오염된 정규모형에서 효율적이고 로버스트하다는 것을 모의실험을 통하여 확인하였다. 또한 세 가지 추정량들에 의해 추정된 모수들을 이용하여 판별하였을 때 자 추정량득의 판별율을 비교함으로써 오염된 정규모형에서 MLE의 대안으로 MGNEDE와 MBWHDE를 사용할 수 있음을 보였다.

  • PDF

Longitudinal Variation of Fish Communities in the Geum River, Korea: Application of the Concept of Beta Diversity and Local Uniqueness

  • Kim, Jeong-Hui;Park, Sang-Hyeon;Baek, Seung-Ho;Hong, Donghyun;Jo, Hyunbin
    • Proceedings of the National Institute of Ecology of the Republic of Korea
    • /
    • 제3권2호
    • /
    • pp.122-128
    • /
    • 2022
  • To present the spatial variation of fish assemblages in the Geum River in Korea, the concept of beta diversity (β-diversity) estimates based on the variance of the community data table was applied. Fish communities and environmental variables were collected from 13 sampling sites along the in mid-low reaches of the River. We calculated the β-diversity and local contribution to beta diversity (LCBD) values at each site depending on the two types of data, 'occurrence' with Jaccard and Sørensen dissimilarity coefficients, and 'abundance' with Hellinger distance. Multivariate and correlation analyses were also performed to determine the relationships between LCBD and other variables, such as community indices and physicochemical and hydrological factors. The β-diversity values of fish communities in the River were estimated as 0.218 and 0.145 for occurrence data table with Jaccard and Sørensen respectively, and 0.268 for abundance data. Similar patterns of LCBD along the sampling sites were detected in two dissimilarity measurements of occurrence table, and LCBD values with abundance data were slightly different. The LCBD values are strongly correlated with community indices, and also suitable for indicating the uniqueness of fish assemblages. However, further research is needed to determine the LCBD value as an indicator of environmental variability.