• The Korean Journal of Applied Statistics
• /
• v.19 no.2
• /
• pp.305-317
• /
• 2006

Outliers adversely affect model identification, parameter estimation, and forecast in time series data. In particular, when outliers consist of a patch of additive outliers, the current outlier detection procedures suffer from the masking and swamping effects which make them inefficient. In this paper, we propose new outlier detection procedure based on high breakdown estimators, called as the dual robust filtering. Empirical and simulation studies in the autoregressive model with orders p show that the proposed procedure is effective.

• 한국경영정보학회:학술대회논문집
• /
• 2007.06a
• /
• pp.619-624
• /
• 2007

본 연구는 유전 알고리듬에 기반한 새로운 도산예측기법을 개발하고 그 기법의 타당성 및 예측 우수성을 검증하는데 목적이 있다. 본 연구에서 제안하는 이진분류기법은 도산기업과 비도산기업을 대표할 수 있는 가상기업(virtual company)을 설정하고, 그 가상기업과 분류대상 기업 간의 유사도를 측정하여 도산여부를 분류하는 방법론으로, 가상기업의 변수 값과 각 변수의 가중치는 훈련용 자료의 분류정확도를 극대화할 수 있도록 유전 알고리듬을 이용하여 구하게 된다. 본 연구에서 제안하는 기법의 타당성을 검증하기 위해 기존의 도산예측기법과 예측성과를 실험을 통해 비교한 결과, 본 연구에서 개발한 기법의 예측력이 기존의 다변량판별분석, 로지스틱 회귀모형, 의사결정나무, 인공신경망 모형보다 높은 수준을 보이는 것을 확인하였다.

• The Korean Journal of Applied Statistics
• /
• v.21 no.5
• /
• pp.825-833
• /
• 2008

Discriminant analysis based on Gaussian mixture models, an useful tool for multi-class classifications, can be extended to semi-supervised learning. We consider a model selection problem for a Gaussian mixture model in semi-supervised learning. More specifically, we adopt Bayesian information criterion to determine the number of subclasses in the mixture model. Through simulations, we illustrate the usefulness of the criterion.

• The Korean Journal of Applied Statistics
• /
• v.18 no.3
• /
• pp.543-553
• /
• 2005

Linear logistic regression is one of the most widely used method for credit scoring in credit risk management. This paper deals with credit scoring using splines based on Logistic regression. Linear splines and an automatic basis selection algorithm are adopted. The final model is an example of the generalized additive model. A simulation using a real data set is used to illustrate the performance of the spline method.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.187-188
• /
• 2019

2018년 12월 20일부터 '제로페이' 시범 서비스를 시작했으며, 2019년 3월 이후 본격적으로 정식 서비스를 시작해 전국으로 확대한다는 방침이다. 이론적 배경으로부터 도출된 가설을 검정하기 위해서 제로페이 사용자를 대상으로 편의표본추출법을 채택하여 설문 조사를 진행할 예정이다. 실증분석에서는 제시된 가설들의 영향 관계를 분석하고 이전 측정항목의 신뢰성과 타당성을 확보하기 위해 신뢰성 분석 및 타당성 분석을 실시한다. 또한 각 변수들 사이의 판별타당성과 변수간의 방향의 정도를 확인하기 위하여 상관관계분석을 실시하고, 경로모형과 가설을 검정한다. 본 연구를 통해 사용자들의 제로페이의 수용에 대한 결정요인을 보다 명확하게 이해할 수 있을 것이며, 관련 연구자나 실무자들은 물론 제로페이 서비스 기업들이 활용할 수 있는 마케팅 전략방안을 제시하고자 한다.

• The korean journal of orthodontics
• /
• v.25 no.2 s.49
• /
• pp.153-163
• /
• 1995

In this study, Receiver Operating Characteristic(ROC) analysis was used to evaluate the ability of cephalometric measurements to identify patients with Class III malocclusions. ROC analysis is the method for determining the validity of a diagnostic measure and for evaluating the relative value of diagnostic tests. The sample consisted of 496 patients with malocclusion. Class III malocclusion is defined as the dental relationship for which The mesiobuccal groove of the lower first molar is deviated mesially from the mesiobuccal cusp of the upper first molar. Of the total sample of 496 patients, 245 had Class III malocclusions. 16 cephalometric measurements were selected, each of which was treated as a diagnostic test. The ROC curves were generated for each cephalometric measurement with intervals of $1.0^{\circ}$ for angular measurements, 1.0mm for linear measurements. The area under the ROC curves was measured for direct comparison among different diagnostic tests. The results were as follows; 1. The 'Wits' appraisal was found to be a better diagnostic criterion for the presence of Class III malocclusion than any other commonly'used cephalometric measurement. 2. AB plane angle, ANB angle, App-Bpp distance, AF-BF distance, APDI, Distance of point A and Pog to N perpendicular, maxillomandibular differential had high diagnostic value. 3. Cephalometric measurements which evaluate the position of the mandible had moderate diagnostic value. 4. Cephalometric measurements related to the maxilla discriminated least between patients with and without Class III malocclusion.

• Journal of Korean Society of Transportation
• /
• v.28 no.5
• /
• pp.17-29
• /
• 2010

The main objective of this study is to consider the necessity of introducing owner expenses in estimating the cost of traffic accidents, as is done in some advanced countries. It is known as "owner cost" in the U.S, "enterprise project price" in Japan, and "human cost" in the U.K. To achieve this, the seriousness and present state of post-traumatic stress disorder (PTSD) of traffic accident offenders, which is generally not observed when traffic accidents occur, are evaluated; countermeasures reflecting personal characteristics are prepared according to degree of affliction. The results could be summarized as follows. First, the present condition and seriousness of PTSD from traffic accidents that has been estimated only within the accidents offenders are evaluated through a survey. Second, a distinction analysis model is developed. 'Insomnia caused by accidents' and 'mental daunting in accident situations or similar ones during driving' are chosen as the variables of the discriminant. Third, as a result of the construction of an Ordered Probit Model to comprehend the seriousness of PTSD, the variables turns out in the order of 'seriousness of the most damaged victim, 'nightmare about the accident', 'insomnia experience', 'mental daunting during driving', 'gender', and 'major driving area'. Finally, the cures reflecting characteristics of gender and region are suggested, which are preferred for degree of seriousness of PTSD. The analysis shows that the method to minimize the damage, such as decline of work efficiency, by moving ahead the rehabilitation of the offenders would be suggested and this could minimize the direct and indirect damages that could happen to business owners for whom the drivers work.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.1C
• /
• pp.123-129
• /
• 2010

In this paper, we analyze the code tracking biases of single and double early-minus-late processing schemes which are widely used code tracking method for global navigation satellite systems. The code tracking bias which results from the distortion in symmetry of correlation values is arisen in the presence of multipath signals. To analyze them, two civil signals which are spreading signals modulated by binary phase shift keying and binary offset carrier are considered.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.97-97
• /
• 2023

최근 재해분야에서 인공신경망(ANN), 기계학습(ML), 딥러닝(DL) 등 AI 기술이 활용성이 점차 증가하고 있으며, 센싱정보와 연계한 시설물 안전관리, 원격탐사와 연계한 재해감시(녹조, 산사태, 산불 등), 수문시계열(수위, 유량 등) 예측, 레이더·위성강수 자료의 보정과 예측, 상하수도 관망누수예측 등 다양한 분야에서 AI 기술이 적용되고 그 활용성이 검증된 바 있다. 본 연구에서는 ML, DL, 물리기반신경망(Pysics-informed Neural Networks, PINNs)을 이용한 다양한 재해분석 사례를 소개하고, 그 활용성과 한계에 대해서 논의하고자 한다. 주요사례로는 (1) SAR영상과 기계학습을 이용한 재해피해지역(울진 산불) 감지, (2) 국가 디지털 정보를 이용한 산사태 위험지역 판별(인제 산사태) (3) 기계학습 및 딥러닝 기법을 이용한 위성강수 자료의 보정·예측 및 유출해석, (4) 수리해석을 위한 수치해석분야에서의 PINNs의 적용성(1차원 Saint-Venant 식 해석) 평가 연구결과를 공유한다. 특히, 자료의 입·출력 자료만으로 학습된 인공신경망 모형 대신 지배방정식(물리방정식)을 만족하도록 강제한 PINNs의 경우, 인공신경망 모형보다 우수한 모의능력을 보여주었으며, 향후 복잡한 수리모델링 등 수치해석분야에서 그 활용가능성이 매우 높을 것으로 판단된다.

• The korean journal of orthodontics
• /
• v.27 no.3 s.62
• /
• pp.359-372
• /
• 1997

This study was conducted to discern differences of craniofacial, dentoalveolar structure and model measurements between sex and between class n openbite group and non-openbite group. The sample consisted of 49 adult patients with class Il malocclusion. 24 linear measurements, 22 angular measurements and 12 ratios were selected in lateral cephalometry. Also, arch width, length, anterior crowding, average molar relation were measured or calculated in diagnostic model. The data were evaluated by t-test and multiple discriminant analysis. The results were as follows, 1. Most linear measurements, with the exception of MnBL and AUDH, were significantly larger in male(p<0.05). but, intermaxillary relations and spatial position of maxilla and mandible relative to cranial base were not different for both sex. 2. With the exception of upper and lower anterior crowding, lower arch width, upper arch length, AMR, male exhibited significantly larger measurements in model analysis (p<0.05). 3. Size differences of maxilla and mandible between openbite and non-openbite group were not significant(p>0.05). but openbite group showed significantly increased genial angle(p<0.05), FH-CoGo(p<0.01), FH-NA(p<0.01) and FH-NB, FH-NPog (p<0.05). 4. ALFH and PUDH were larger(p<0.05) in openbite group. this result served as compensation for the spatial position of mandible relative to cranial base. AUPUDH (p<0.001) and ALPLDH(p<0.05) were lower in openbite group. upper anterior crowding was the only measurement which showed difference between openbite and non-openbite group(p<0.05). 5. For the purpose of classifying adult class n openbite and non-openbite group, multiple discriminant analysis was done genial angle, ALPLDH, AUPUDH, FH-NA were included in multiple discriminant equation. 39 cases($92.86\%$) were correctly classified when applied to the sample used in this study.