• The Korean Journal of Applied Statistics
• /
• v.24 no.2
• /
• pp.269-278
• /
• 2011

There are many researches that have considered the distribution functions and appropriate covariates corresponding to the scores in order to improve the accuracy of a diagnostic test, including the ROC curve that is represented with the relations of the sensitivity and the specificity. The ROC analysis was used by the regression model including some covariates under the assumptions that its distribution function is known or estimable. In this work, we consider a general situation that both the distribution function and the elects of covariates are unknown. For the ROC analysis, the mixtures of normal distributions are used to estimate the distribution function fitted to the credit evaluation data that is consisted of the score random variable and two sub-populations of parameters. The AUC measure is explored to compare with the nonparametric and empirical ROC curve. We conclude that the method using normal mixtures is fitted to the classical one better than other methods.

• The Korean Journal of Applied Statistics
• /
• v.35 no.1
• /
• pp.35-47
• /
• 2022

The receiver operating characteristic (ROC) curve was developed to quantify the classification ability of marker values (covariates) on the response variable and has been extended to survival data with diverse missing data structure. When survival data is understood as binary data (status of being alive or dead) at each time point, the ROC curve expressed at every time point results in time-dependent ROC curve and time-dependent area under curve (AUC). In particular, a follow-up study brings the change of cohort and incomplete data structures such as censoring and competing risk. In this paper, we review time-dependent ROC estimators under several contexts and perform simulation to check the performance of each estimators. We analyzed a dementia dataset to compare the prognostic power of markers.

• Communications for Statistical Applications and Methods
• /
• v.19 no.2
• /
• pp.277-286
• /
• 2012

For credit assessment models, the ROC curves evaluate the classification performance using two univariate cumulative distribution functions of the false positive rate and true positive rate. In this paper, it is extended to two bivariate normal distribution functions of default and non-default borrowers; in addition, the bivariate ROC curves are proposed to represent the joint cumulative distribution functions by making use of the linear function that passes though the mean vectors of two score random variables. We explore the classification performance based on these ROC curves obtained from various bivariate normal distributions, and analyze with the corresponding AUROC. The optimal threshold could be derived from the bivariate ROC curve using many well known classification criteria and it is possible to establish an optimal cut-off criteria of bivariate mixture distribution functions.

• Journal of radiological science and technology
• /
• v.40 no.2
• /
• pp.237-243
• /
• 2017

This study evaluated the applicability of computer-aided diagnosis by retrospective analysis of GLCM algorithm based on cytopathological diagnosis of normal and malignant nodules in thyroid ultrasound images. In the experiment, the recognition rate and ROC curve of thyroid malignant nodule were analyzed using 6 parameters of GLCM algorithm. Experimental results showed 97% energy, 93% contrast, 92% correlation, 92% homogeneity, 100% entropy and 100% variance. Statistical analysis showed that the area under the curve of each parameter was more than 0.947 (p = 0.001) in the ROC curve, which was significant in the recognition of thyroid malignant nodules. In the GLCM, the cut-off value of each parameter can be used to predict the disease through analysis of quantitative computer-aided diagnosis.

• The Korean Journal of Applied Statistics
• /
• v.24 no.1
• /
• pp.185-191
• /
• 2011

In the credit evaluation study with the assumption of mixture distributions, the ROC curve is a useful method to explore the discriminatory power of default and non-default borrowers. The AROC curve is an adjusted ROC curve that can be identified with the corresponding score and is mathematically analyzed in this work. We obtain patterns of this curve by applying normal distributions. Moreover, the relationship between the AROC curve and many classification accuracy statistics are explored to find the optimal threshold. In the case of equivalent variances of two distributions, we obtain that the local minimum of the AROC curve is estimated at the optimal threshold to maximize certain classification accuracies.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.367-367
• /
• 2019

최근 기후변화로 인해 국지성 호우 및 태풍의 빈도가 빈발하고 및 규모가 커지고 있으며 그로 인한 홍수피해규모는 증가하고 있다. 본 논문에서는 도시 지역의 호우로 인한 침수유발 강우량을 산정하는 기법의 정확도를 산정하는데 목적이 있으며 이를 위해 ROC(Receiver Operation Characteristic Curve) 분석을 이용하였다. 본 논문에서는 분포형 홍수해석 모형인 S-RAT 모형과 2차원 침수해석 모형 FLO-2D을 커플링하여 호우로 인한 침수해석을 실시하였으며 강우시나리오는 설계 강우 200mm의 강우를 10% 간격으로 증가시켜 강우량 대비 침수심 자료를 모의하였다. 모의한 침수심 자료를 이용하여 유역 격자를 $1km{\times}1km$ 별 강우량-침수심 관계곡선식을 제시하였으며 개발된 곡선식을 이용하여 특정 침수심(20cm)을 유발시키는 강우량(한계강우량)을 산정하였다. 정확도 산정은 ROC(Receiver Operation Characteristic Curve) 분석 방법을 이용하여 침수 유무의 적중률에 따른 민감도와 특이도를 이용하여 AUC(Area Under the Curve)의 점수로 정확도를 판단하였다. 본 논문에서는 본 논문에서 제시한 한계강우량의 정확도를 판단하기 위하여 2011년 7월의 사당역 일대 침수사례를 이용하였다. 실제 침수정보가 없어 실제 호우사상과 실제 하수관망을 고려할 수 있는 SWMM 모형을 이용하여 침수분석을 실시하였다. 분석 결과 평균 ROC는 약 0.7로 나타났으며 5 단계의 구분에서 Fair 단계로 적정 수준의 정확도를 확보한 것으로 나타났다.

• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.27 no.2
• /
• pp.7-13
• /
• 1997

조직학적으로 유용성이 입증된 산탈회법을 이용한 인접면 비교적 초기 치아 우식의 병소를 형성하여 진단율을 조사하였다. 산 용액을 이용하여 20개 인접면 치아우식을 20개 소구치에 형성하였고, 37개 인접면 치아우식을 30개 대구치에 형성하였다. 건전한 소구치 20개, 대구치 30개를 포함하여 총 96개 치아를 4개씩 나누어 24개의 블록을 형성하였고, 각각 2개 블록의 교합면을 교합시켜서, 교익촬영을 하였다. 촬영 결과를 36명의 치과의사들이인접면 치아우식의 유무를 기록하고, 동시에 및 ROC 분석을 위한 5 개 범주의 판독 기준으로 판독하여 기록하였다. 인접면 치아우식증 유, 무만으로 판독한 결과 진단의 sensitivity는 0.71, specificity는 0.78 이였다. ROC 분석 한 결과의 곡선도표 아래부분의 평균 면적은 약 0.806 이였다. 치아우식증 유무만으로 진단한 결과는 특정한 sensitivity와 specificity 만을 나타내지만, ROC 분석 결과는 주관적 진단 기준과 구별되는 고유의 진단 능력을 표시하는 1-specificity(False Positive)의 변화에 따른 sensitivity(True Positive)의 변화를 연속적으로 나타내어 주었다.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.6
• /
• pp.1367-1375
• /
• 2015

Consider the linear growth models for longitudinal data analysis. Several kind of linear growth models are selected such as time-effect and random-effect models as well as a dummy variable included model. In this work, simulation data are generated with normality assumption, and both binormal ROC curve and AUC are obtained and compared for various linear growth models. It is found that ROC curves have different shapes and AUC increase slowly, as values of the covariance increase and the time passes for random-effect models. On the other hand, AUC increases very fast as values of covariance decrease. When the covariance has positive value, we explored that the variances of random-effect models increase and the increment of AUC is smaller than that of AUC for time-effect models. And the increment of AUC for time-effect models is larger than the increment for random-effect models.

• The Korean Journal of Applied Statistics
• /
• v.27 no.5
• /
• pp.773-786
• /
• 2014

We propose the multiplication of false rates (MFR) which is a classification accuracy criteria and an area type of rectangle from ROC curve. Optimal threshold obtained using MFR is compared with other criteria in terms of classification performance. Their optimal thresholds for various distribution functions are also found; consequently, some properties and advantages of MFR are discussed by comparing FNR and FPR corresponding to optimal thresholds. Based on general cost function, cost ratios of optimal thresholds are computed using various classification criteria. The cost ratios for cost curves are observed so that the advantages of MFR are explored. Furthermore, the de nition of MFR is extended to multi-dimensional ROC analysis and the relations of classification criteria are also discussed.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.4
• /
• pp.669-680
• /
• 2017

This study evaluated the consistency of the standard flow to forecast low-flow based on various drought indices. The data used in this study were streamflow data at the Gurye2 station located in the Seomjin River and the Angang station located in the Hyeongsan River, as well as rainfall data of nearby weather stations (Namwon and Pohang). Using streamflow data, the streamflow accumulation drought index (SADI) was developed in this study to represent the hydrological drought condition. For SADI calculations, the threshold of drought was determined by a Change-Point analysis of the flow pattern and a reduction factor was estimated based on the kernel density function. Standardized runoff index (SRI) and standardized precipitation index (SPI) were also calculated to compared with the SADI. SRI and SPI were calculated for the 30-, 90-, 180-, and 270-day period and then an ROC curve analysis was performed to determine the appropriate time-period which has the highest consistency with the standard flow. The result of ROC curve analysis indicated that for the Seomjin River-Gurye2 station SADI_C3, SRI30, SADI_C1, SADI_C2, and SPI90 were confirmed in oder of having high consistency with standard flow under the attention stage and for the Hyeongsan River-Angang station, SADI_C3, SADI_C1, SPI270, SRI30, and SADI_C2 have order of high consistency with standard flow under the attention stage.