• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.39 no.6
• /
• pp.667-673
• /
• 2019

Global warming may accelerate climate change and may increase disaster caused by strong winds. This research studied a method for a nonstationary frequency analysis considering the linear trend over time. The Bayesian method was used to estimate the posterior distribution of the parameters for the extreme value distribution of the annual maximum wind speed at Jeju Airport. The nonstationary frequency analysis was performed based on the Monte Carlo Markov Chain simulation and the Gibbs sampling. The estimated wind speeds by nonstationary frequency analysis was larger than those by stationary analysis. The conventional frequency analysis procedure assuming stationarity is likely to underestimate the future design wind speed in the region where statistically significant trend exists.

• The Korean Journal of Applied Statistics
• /
• v.30 no.6
• /
• pp.1015-1025
• /
• 2017

Bayesian personalized ranking (BPR) is a state-of-the-art recommendation system techniques for implicit feedback data. Unfortunately, there might be a loss of information because the BPR model considers only the binary transformation of implicit feedback that is non-binary data in most cases. We propose a modified BPR method using a level of confidence based on the size or strength of implicit feedback to overcome this limitation. The proposed method is useful because it still has a structure of interpretable models for underlying personalized ranking i.e., personal pairwise preferences as in the BPR and that it is capable to reflect a numerical size or the strength of implicit feedback. We propose a computation algorithm based on stochastic gradient descent for the numerical implementation of our proposal. Furthermore, we also show the usefulness of our proposed method compared to ordinary BPR via an analysis of steam video games data.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.34 no.6
• /
• pp.1787-1796
• /
• 2014

The main purpose of this study is to assess the sensitivity of meteorological drought indices in probabilistic perspective using Bayesian Network model. In other words, this study analyzed interrelationships between various drought indices and investigated the order of the incident. In this study, a Bayesian Network model was developed to evaluate meteorological drought characteristics by employing the percent of normal precipitation (PN) and Standardized Precipitation Index (SPI) with various time scales such as 30, 60, and 90 days. The sensitivity analysis was also performed for posterior probability of drought indices with various time scales. As a result, this study found out interdependent relationships among various drought indices and proposed the effective application method of SPI to drought monitoring.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.451-451
• /
• 2015

최근 1900년부터 1990년 사이 해수면은 매년 평균 1.2mm 상승했지만 1990년부터는 매년 평균 3mm씩 높아지고 있으며, 이에 1990년부터 현재까지 해수면 수위의 상승속도가 이전 90년 동안 측정된 수치보다 2.5배 빠르다는 연구결과가 발표되었다. 해수면 상승으로 인한 피해는 범람과 침식을 야기할 수 있으며 해일 및 폭풍으로 인한 피해를 증가시킴으로 물질적 피해와 인명 피해를 유발할 수 있다. 이러한 이유로 해수면 상승에 따른 과학적인 분석과 신뢰성 있는 전망을 통하여 해수면 상승에 따른 대응과 대비가 필요하다. 이에 본 연구에서는 비정상성 빈도해석 방법을 통하여 미래의 해수면 상승을 고려할 수 있는 비정상성 빈도해석 기법을 개발하였다. 본 연구에서는 극치사상을 추출하기 위해 국립해양조사원 (Korea Hydrographic and Oceanographic Administration, KHOA)에서 관리한 45개 조위관측소의 시 조위 자료를 이용하였다. 45개 조위관측소의 한 시간 단위 자료로부터 연최대 및 연평균 조위계열 (annual average and annual maximum sea level series)을 추출하였다. 본 연구에서는 한반도 해안을 동해안, 서해안, 남해안, 제주 권역으로 구분하고 빈도 해석의 신뢰성을 만족하기 위해 자료 구축기간이 20년 이상이며, 각 해안을 나타낼 수 있는 지점을 선정하였다. 비정상성 빈도해석은 Gumbel 극치분포를 적용하였으며, 계층적 Bayesian 기법을 결합하여 매개변수들에 대한 사후분포를 추정하였다. 본 연구에서는 대부분의 지점에서 비정상성 빈도해석 결과와 정상성 빈도해석 결과와 상당한 차이를 보여주고 있으며, 이는 주로 정상성 가정에 기인하는 문제점으로 판단된다. 향후 기후변화에 따른 연안지역의 홍수 및 사회기반시설의 위험도를 평가하기 위해서는 비정상성을 고려한 빈도해석 절차의 수립과 적용이 필요할 것으로 판단된다.

• The Korean Journal of Applied Statistics
• /
• v.32 no.4
• /
• pp.619-630
• /
• 2019

Dyslipidemia is a representative chronic disease affecting Koreans that requires continuous management. It is also a known risk factor for cardiovascular disease such as hypertension and diabetes. However, it is difficult to diagnose vascular disease without a medical examination. This study identifies risk factors for the recognition and prevention of dyslipidemia. By integrating them, we construct a statistical instrumental nomogram that can predict the incidence rate while visualizing. Data were from the Korean National Health and Nutrition Examination Survey (KNHANES) for 2013-2016. First, a chi-squared test identified twelve risk factors of dyslipidemia. We used a naïve Bayesian classifier model to construct a nomogram for the dyslipidemia. The constructed nomogram was verified using a receiver operating characteristics curve and calibration plot. Finally, we compared the logistic nomogram previously presented with the Bayesian nomogram proposed in this study.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.1
• /
• pp.29-37
• /
• 2017

It has gradually become important to estimate a forest stand volume utilizing LiDAR data. Recently, various statistical models including a linear regression model has been introduced to estimate a forest stand volume using LiDAR data. One of limitations of the current approaches is in that the accuracy of observed forest stand volume data, which is used as a response variable, is questionable unstable. To overcome this limitation, we consider a spatial structure for a forest stand volume. In this research, we propose a hierarchical model for applying a spatial structure to a forest stand volume. The proposed model is applied to the LiDAR data and the forest stand volume for Bonghwa, Gyeongsangbuk-do.

• The Korean Journal of Applied Statistics
• /
• v.21 no.4
• /
• pp.705-716
• /
• 2008

The objectives of this study are the detection and forecast of climate change signal in the annual mean of surface temperature data, which are generated by MRI/JMA CGCM over the Korean Peninsula. MRI/JMA CGCM outputs consist of control run data(experiment with no change of $CO_2$ concentration) and scenario run data($CO_2$ 1%/year increase experiment to quadrupling) during 142 years for surface temperature and precipitation. And ECMWF reanalysis data during 43 years are used as observations. All data have the same spatial structure which consists of 42 grid points. Two statistical models, the Bayesian fingerprint method and the regression model with autoregressive error(AUTOREG model), are separately applied to detect the climate change signal. The forecasts up to 2100 are generated by the estimated AUTOREG model only for detected grid points.

• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.407-426
• /
• 2003

We consider the missing covariates problem in generalized estimating equations(GEE) model. If the covariate is partially missing, GEE can not be calculated. In this paper, we study the performance of 7 imputation methods to handle missing covariates in GEE models, and the properties of GEE estimators are investigated after missing covariates are imputed for ordinal data of repeated measurements. The 7 imputation methods include i) Naive Deletion ii) Sample Average Imputation iii) Row Average Imputation iv) Cross-wave Regression Imputation v) Carry-over Imputation vi) Bayesian Bootstrap vii) Approximate Bayesian Bootstrap. A Monte-Carlo simulation is used to compare the performance of these methods. For the missing mechanism generating the missing data, we assume ignorable nonresponse. Furthermore, we generate missing covariates with or without considering wave nonresp onse patterns.

• Journal of the Korean Society for Railway
• /
• v.12 no.5
• /
• pp.656-662
• /
• 2009

In general, the analysis of travelers' mode choice behavior is accomplished by developing the utility functions which reflect individual's preference of mode choice according to their demographic and travel characteristics. In this paper, we propose a methodology that takes the spatial effects of individuals' departure locations into account in the mode choice model. The statistical models considered here are spatial logistic regression model and conditional autoregressive model taking a spatial association parameter into account. We employed the Bayesian approach in order to obtain more reliable parameter estimates. The proposed methodology allows us to estimate mode shares by departure places even though the survey does not cover all areas.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.96-96
• /
• 2021

수문기상인자 중 하나인 증발산량은 수자원 계획 및 관리 시 고려되며, 특히 물수지 모형 등의 입력자료로 활용된다. 우리나라를 포함한 각국 기상청 및 국제기구에서는 직접 관측이 아닌 FAO56 Penman-Monteith(PM)을 통해 증발산량을 산출하고 있다. FAO56 PM 방법은 복사(radiation), 대기온도(air temperature), 습도(humidity), 풍속(wind speed) 등의 기상인자로부터 기준증발산량(reference evapotransipiration)을 추정하며, 상대적으로 높은 정확성을 보여준다. 그러나 FAO56 PM 방법은 많은 기상인자를 요구하므로 미계측 유역을 포함한 일부지역에 대한 증발산량 자료 구축이 어려운 실정이다. 또한, 기준증발산량의 특성이 시간에 따라 변화하므로 비정상성(nonstationary)을 고려한 분석이 요구된다. 본 연구에서는 온도인자 기반의 대체모형(surrogate model)을 개발하여 기준증발산량의 비정상성을 고려하고자 한다. 한강유역에 위치한 관측소를 대상으로 모형을 개발하였으며, 시간에 따라 변동하는 기준증발산량의 특성을 고려하기 위해 Bayesian 추론기법을 통해 매개변수를 시간에 따라 추정하였다. 또한, 본 연구에서는 대체모형으로 산정된 증발산량을 활용해 가뭄지수인 EDDI(evaporative demand drought index)를 제시하였다. 가뭄 모니터링 및 조기 경보 안내를 위해 개발된 EDDI를 활용하여 기존 가뭄보다 빠르게 진행되는 초단기 가뭄(flash drought)를 평가하였다. 본 연구에서 개발된 모형은 미계측 지역에서도 적용이 가능하므로 수자원분야에서 활용성이 높을 것으로 사료된다.