• The Bulletin of The Korean Astronomical Society
• /
• v.41 no.1
• /
• pp.80.1-80.1
• /
• 2016

There are probabilistic forecast models for solar flare occurrence, which can be evaluated by various skill scores (e.g. accuracy, critical success index, heidek skill score, and true skill score). Since these skill scores assume that two types of forecast errors (i.e. false alarm and miss) are equal or constant, which does not take into account different situations of users, they may be unrealistic. In this study, we make an evaluation of a probabilistic flare forecast model [Lee et al., 2012] which use sunspot groups and its area changes as a proxy of flux emergence. We calculate daily solar flare probabilities from 2011 to 2014 using this model. The skill scores are computed through contingency tables as a function of forecast probability, which corresponds to the maximum skill score depending on flare class and type of a skill score. We use a value score with cost/loss ratio, relative importance between the two types of forecast errors. The forecast probability (y) is linearly changed with the cost/loss ratio (x) in the form of y=ax+b: a=0.88; b=0 (C), a=1.2; b=-0.05(M), a=1.29; b=-0.02(X). We find that the forecast model has an effective range of cost/loss ratio for each class flare: 0.536-0.853(C), 0.147-0.334(M), and 0.023-0.072(X). We expect that this study would provide a guideline to determine the probability threshold and the cost/loss ratio for space weather forecast.

• The Korean Journal of Applied Statistics
• /
• v.22 no.2
• /
• pp.237-248
• /
• 2009

The probability of default on the credit evaluation study is represented as a linear combination of two distributions of default and non-default, and the distribution of the probability of default are generally known in most cases. Except the well-known Kolmogorov-Smirnov statistic for testing the identity of two distribution, Kuiper, Cramer-Von Mises, Anderson-Darling, and Watson test statistics are introduced in this work. Under the assumption that the population distribution is known, modified Cramer-Von Mises, Anderson-Darling, and Watson statistics are proposed. Based on score data generated from various probability density functions of the probability of default, the modified test statistics are discussed and compared.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.12
• /
• pp.6009-6027
• /
• 2019

Support vector machine (SVM) classifiers have been widely used for object detection. These methods usually locate the object by finding the region with maximal score in an image. With bag-of-features representation, the SVM score of an image region can be written as the sum of its inside feature-weights. As a result, the searching process can be executed efficiently by using strategies such as branch-and-bound. However, the feature-weight derived by optimizing region classification cannot really reveal the category knowledge of a feature-point, which could cause bad localization. In this paper, we represent a region in an image by a collection of local feature-points and determine the object by the region with the maximum posterior probability of belonging to the object class. Based on the Bayes' theorem and Naive-Bayes assumptions, the posterior probability is reformulated as the sum of feature-scores. The feature-score is manifested in the form of the logarithm of a probability ratio. Instead of estimating the numerator and denominator probabilities separately, we readily employ the density ratio estimation techniques directly, and overcome the above limitation. Experiments on a car dataset and PASCAL VOC 2007 dataset validated the effectiveness of our method compared to the baselines. In addition, the performance can be further improved by taking advantage of the recently developed deep convolutional neural network features.

• The Bulletin of The Korean Astronomical Society
• /
• v.40 no.1
• /
• pp.84.2-84.2
• /
• 2015

There are probabilistic forecast models for solar flare occurrence, which can be evaluated by various skill scores (e.g. accuracy, critical success index, heidek skill score, true skill score). Since these skill scores assume that two types of forecast errors (i.e. false alarm and miss) are equal or constant, which does not take into account different situations of users, they may be unrealistic. In this study, we make an evaluation of a probabilistic flare forecast model (Lee et al. 2012) which use sunspot groups and its area changes as a proxy of flux emergence. We calculate daily solar flare probabilities from 1996 to 2014 using this model. Overall frequencies are 61.08% (C), 22.83% (M), and 5.44% (X). The maximum probabilities computed by the model are 99.9% (C), 89.39% (M), and 25.45% (X), respectively. The skill scores are computed through contingency tables as a function of forecast probability, which corresponds to the maximum skill score depending on flare class and type of a skill score. For the critical success index widely used, the probability threshold values for contingency tables are 25% (C), 20% (M), and 4% (X). We use a value score with cost/loss ratio, relative importance between the two types of forecast errors. We find that the forecast model has an effective range of cost/loss ratio for each class flare: 0.15-0.83(C), 0.11-0.51(M), and 0.04-0.17(X), also depending on a lifetime of satellite. We expect that this study would provide a guideline to determine the probability threshold for space weather forecast.

• The Korean Journal of Applied Statistics
• /
• v.36 no.4
• /
• pp.323-335
• /
• 2023

In causal analysis of high dimensional data, it is important to reduce the dimension of covariates and transform them appropriately to control confounders that affect treatment and potential outcomes. The augmented inverse probability weighting (AIPW) method is mainly used for estimation of average treatment effect (ATE). AIPW estimator can be obtained by using estimated propensity score and outcome model. ATE estimator can be inconsistent or have large asymptotic variance when using estimated propensity score and outcome model obtained by parametric methods that includes all covariates, especially for high dimensional data. For this reason, an ATE estimation using an appropriate dimension reduction method and semiparametric model for high dimensional data is attracting attention. Semiparametric method or sparse sufficient dimensionality reduction method can be uesd for dimension reduction for the estimation of propensity score and outcome model. Recently, another method has been proposed that does not use propensity score and outcome regression. After reducing dimension of covariates, ATE estimation can be performed using matching. Among the studies on ATE estimation methods for high dimensional data, four recently proposed studies will be introduced, and how to interpret the estimated ATE will be discussed.

• The Korean Journal of Applied Statistics
• /
• v.36 no.5
• /
• pp.381-397
• /
• 2023

The most representative design used in clinical trials is randomization, which is used to accurately estimate the treatment effect. However, comparison between the treatment group and the control group in an observational study without randomization is biased due to various unadjusted differences, such as characteristics between patients. Propensity score weighting is a widely used method to address these problems and to minimize bias by adjusting those confounding and assess treatment effects. Inverse probability weighting, the most popular method, assigns weights that are proportional to the inverse of the conditional probability of receiving a specific treatment assignment, given observed covariates. However, this method is often suffered by extreme propensity scores, resulting in biased estimates and excessive variance. Several alternative methods including trimming, overlap weights, and matching weights have been proposed to mitigate these issues. In this paper, we conduct a simulation study to compare performance of various propensity score weighting methods under diverse situation, such as limited overlap, misspecified propensity score, and treatment contrary to prediction. From the simulation results overlap weights and matching weights consistently outperform inverse probability weighting and trimming in terms of bias, root mean squared error and coverage probability.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.73-75
• /
• 2022

Scene Text Recognition is a technology used in the field of artificial intelligence that requires manless robot, automatic vehicles and human-computer interaction. Though scene text images are distorted by noise interference, such as illumination, low resolution and blurring. Unlike previous studies that recognized only English, this paper shows a strong recognition accuracy including various characters, English, Korean, special character and numbers. Instead of selecting only one class having the highest probability value, a candidate word can be generated by considering the probability value of the second rank as well, thus a method can be corrected an existing language misrecognition problem.

• Journal of Korean Neurosurgical Society
• /
• v.41 no.4
• /
• pp.217-223
• /
• 2007

Objective : The apparent increase in the incidence of the intracranial meningiomas in the elderly is due in part to improved diagnostic tools and improved span of life. The authors carried out a retrospect study to validate the use of the Clinical-Radiological Grading System [CRGS] as a clinical tool to orientate surgical decision making in elderly patients and to explore prognostic factors of survival. Methods : From January 1997 to January 2006, the authors consecutively recruited and surgically treated 20 patients older than 65 years of age with radiologic findings of intracranial meningiomas and a preoperative evaluation based on the CRGS. Results : High CRGS score was associated with a higher probability of good outcome [p=0.004] and a lower probability of postoperative complications [p=0.049]. Among the different subset items of the CRGS score, larger maximum tumor diameters [$D{\geqq}4cm$] and the presence of a severe peritumoral edema were associated with incidence rate of postoperative poor outcome and complications [p<0.05]. Additionally, the critical location of the tumor was also correlated with poor outcome [p<0.05]. Conclusion : A CRGS score higher than 13 is a good prognostic indication of survival. The CRGS score is a useful and practical tool for the selection of elderly patients affected by intracranial meningiomas as surgical candidates.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2008.05a
• /
• pp.1027-1031
• /
• 2008

본 연구는 한반도 영역을 대상으로 2001년 7, 8월과 2002년 6월로 홍수기를 대상으로 RDAPS 모형, AWS, 상층기상관측(upper-air sounding)의 자료를 이용하였다. 또한 수치예보자료를 범주적 예측확률로 변환하고 인공신경망기법(ANN)을 이용하여 강수발생확률의 예측정확성을 향상시키는데 있다. 신경망의 예측인자로 사용된 대기변수는 500/ 750/ 1000hpa에서의 지위고도, 500-1000hpa에서의 층후(thickness), 500hpa에서의 X와 Y의 바람성분, 750hpa에서의 X와 Y의 바람성분, 표면풍속, 500/ 750hpa/ 표면에서의 온도, 평균해면기압, 3시간 누적 강수, AWS관측소에서 관측된 RDAPS모형 실행전의 6시간과 12시간동안의 누적강수, 가강수량, 상대습도이며, 예측변수로는 강수발생확률로 선택하였다. 강우는 다양한 대기변수들의 비선형 조합으로 발생되기 때문에 예측인자와 예측변수 사이의 복잡한 비선형성을 고려하는데 유용한 인공신경망을 사용하였다. 신경망의 구조는 전방향 다층퍼셉트론으로 구성하였으며 역전파알고리즘을 학습방법으로 사용하였다. 강수예측성과의 질을 평가하기 위해서 $2{\times}2$ 분할표를 이용하여 Hit rate, Threat score, Probability of detection, Kuipers Skill Score를 사용하였으며, 신경망 학습후의 강수발생확률은 학습전의 강수발생확률에 비하여 한반도영역에서 평균적으로 Kuipers Skill Score가 0.2231에서 0.4293로 92.39% 상승하였다.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.22 no.1
• /
• pp.1-10
• /
• 2019

In this paper, we propose the track initiation algorithm based on the weighted score for TWS radar tracking. This algorithm utilizes radar velocity information to calculate the probabilistic track score and applies the Non-Maximum-Suppression(NMS) to confirm the targets to track. This approach is understood as a modification of a conventional track initiation algorithm in a probabilistic manner. Also, we additionally apply the weighted Hough transform to compensate a measurement error, and it helps to improve the track detection probability. We designed the simulator in order to demonstrate the performance of the proposed track initiation algorithm. The simulation result show that the proposed algorithm, which reduces about 40 % of a false track probability, is better than the conventional algorithm.