정규혼합에서 분류정확도 측도들의 최적기준 = Optimal Criterion of Classification Accuracy Measures for Normal Mixture

국문 초록 (Abstract)

두 분포함수의 혼합모형을 가정한 자료에서 적절한 분류점을 찾고 평가하는 것은 중요한 문제이다. 분류정확도 측도로 많이 사용하는 아홉 종류의 MVD, Youden지수, (0,1)까지 최단기준, 수정된(0,1)까지 최단 기준, SSS, 대칭점, 정확도면적, TA, TR에 대하여 설명하고, 이 측도들의 관계를 발견하면서 정확도 측도들의 조건을 몇 개의 범주로 군집화한다. 정규혼합분포를 가정하여 군집된 측도들에 기반하는 분류점들을 구하고, 그 분류점에 대응하는 제I종 오류율과 제II종 오류율 그리고 두 종류의 오류율합을 구하여 크기를 비교하고 토론하다. 추정된 혼합분포에 대하여 어떤 분류 정확도 측도의 제I종과 II종 오류율 또는 오류율합이 최소인지를 탐색할 수 있으며 자주 인용하는 정확도 측도의 장점과 단점을 파악할 수 있다.

다국어 초록 (Multilingual Abstract)

For a data with the assumption of the mixture distribution, it is important to find an appropriate threshold and evaluate its performance. The relationship is found of well-known nine classification accuracy measures such as MVD, Youden's index, the closest-to-(0, 1) criterion, the amended closest-to-(0, 1) criterion, SSS, symmetry point, accuracy area, TA, TR. Then some conditions of these measures are categorized into seven groups. Under the normal mixture assumption, we calculate thresholds based on these measures and obtain the corresponding type I and II errors. We could explore that which classification measure has minimum type I and II errors for estimated mixture distribution to understand the strength and weakness of these classification measures.

참고문헌 (Reference)

1 홍종선, "혼합분포에서 최적분류점" 한국통계학회 23 (23): 13-28, 2010

2 Tasche, D., "Validation of internal rating systems and PD estimates"

3 Finley, J. P, "Tornado predictions" 1 : 85-88, 1884

4 Perkins, N. J, "The inconsistency of "optimal" cutpoints obtained using two criteria based on the receiver operating characteristic curve" 163 : 670-675, 2006

5 Ward, C. D, "The differential positive rate, a derivative of receiver operating characteristic curves useful in comparing tests and determining decision levels" 32 : 1428-1429, 1986

6 Pepe, M. S, "The Statistical Evaluation of Medical Tests for Classification and Prediction" University Press 2003

7 Zhou, X. H., "Statistical Methods in Diagnostic Medicine" Wiley 2002

8 Provost, F, "Robust classification for imprecise environments" 42 : 203-231, 2001

9 홍종선, "ROC와 CAP 곡선에서의 최적 분류점" 한국통계학회 22 (22): 911-921, 2009

10 Fawcett, T., "ROC graphs: notes and practical considerations for data mining researchers" HP Laboratories 2003

11 Krzanowski, W. J, "ROC Curves for Continuous Data" Champman & Hall/CRC 2009

12 Feinstein, A. R., "Principles of Medical Statistics" Chapman & Hall/CRC 2002

13 Engelmann, B., "Measuring the discriminative power of rating systems, series 2: Banking and Financial Supervision, 01"

14 Liu, C., "Measuring the accuracy of species distribution models: a review" 2009

15 Sobehart, J. R, "Measuring default accurately" 14 : 31-33, 2001

16 Connell, F. A, "Measures of gain in certainty from a diagnostic test" 121 : 744-753, 1985

17 Youden, W. J., "Index for rating diagnostic test" 3 : 32-35, 1950

18 Greiner, M. M, "Epidemiologic issues in the validation of veterinary diagnostic tests" 45 : 3-22, 2000

19 Brasil,P, "Diagnostic test accuracy evaluation for medical professionals, package DiagnosisMed in R"

20 Cantor, S. B, "Determining the area under the ROC curve for a binary diagnostic test" 20 : 468-470, 2000

21 Moses, L. E., "Combining independent studies of a diagnostic test into a summary ROC curve: Data-analytic approaches and some additional considerations" 12 : 1293-1316, 1993

22 Lambert, J, "A macro for getting more out of your ROC curve" 231-, 2008

23 Freeman, E. A, "A comparison of the performance of threshold criteria for binary classification in terms of predicted prevalence and kappa" 217 : 48-58, 2008

24 Cantor, S. B., "A comparison of C/B ratios from studies using receiver operating characteristic curve analysis" 52 : 885-892, 1999

25 Velez, D. R., "A balanced accuracy function for epistasis modeling in imbalanced datasets using multifactor dimensionality reduction" 31 : 306-315, 2007