• 한국해안·해양공학회논문집
• /
• 제34권4호
• /
• pp.109-118
• /
• 2022

정점 표층 수온 관측 데이터 중 결측 구간의 데이터를 양방향 순환신경망(Bidirectional Recurrent Neural Network, BiRNN) 기법을 이용하여 보간하였다. 인공지능 기법 중 시계열 데이터에 일반적으로 활용되는 Recurrent Neural Networks(RNNs)은 결측 추정 위치까지의 시간 흐름 방향 또는 역방향으로만 추정하기 때문에 장기 결측 구간에는 추정 성능이 떨어진다. 반면, 본 연구에서는 결측 구간 전후의 양방향으로 추정을 하여 장기 결측 데이터에 대해서도 추정 성능을 높일 수 있다. 또한 관측점 주위의 가용한 모든 데이터(수온, 기온, 바람장, 기압, 습도)를 사용함으로써, 이들 상관관계로부터 보간 데이터를 함께 추정하도록 하여 보간 성능을 더욱 높이고자 하였다. 성능 검증을 위하여 통계 기반 모델인 Multivariate Imputation by Chained Equations(MICE)와 기계학습 기반의 Random Forest 모델, 그리고 Long Short-Term Memory(LSTM)을 이용한 RNN 모델과 비교하였다. 7일간의 장기 결측에 대한 보간에 대해서 BiRNN/통계 모델들의 평균 정확도가 각각 70.8%/61.2%이며 평균 오차가 각각 0.28도/0.44도로 BiRNN 모델이 다른 모델보다 좋은 성능을 보인다. 결측 패턴을 나타내는 temporal decay factor를 적용함으로써 BiRNN 기법이 결측 구간이 길어질수록 보간 성능이 기존 방법보다 우수한 것으로 판단된다.

• 한국방재학회 논문집
• /
• 제1권2호
• /
• pp.115-121
• /
• 2001

실측자료가 충분하지 못한 단기간의 유출량 자료로부터 추계학적 모형에 의해 장기간의 자료를 모의발생시키는 목적은 수공구조물의 설계에 필요한 설계홍수량의 산정 및 수자원 시스템의 운영조작 방침을 결정하기 위한 풍부한 입력자료를 제공하는데 있다. 특히 본 연구에서는 단일지점이 아닌 다지점에 대한 지점간 서로의 연관성을 고려한 하천유량의 추계학적인 모의 발생기법인 다변량 자기회귀 모형을 적용하고자 한다. 따라서 본 연구에서는 낙동강유역의 2개 지점에 대하여 다변량 모형을 적용하여 모의 발생된 월유량과 실측치를 통계적으로 비교, 분석하였다. 모의발생된 월유량과 실측치를 평균, 분산, 왜곡도, 상관관계 등에 의해 비교, 분석한 결과 모의발생된 월유량과 실측치는 통계적으로 매우 유사하게 나타났다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2022년도 학술발표회
• /
• pp.150-150
• /
• 2022

This study investigates the roles of decomposition methods on high accuracy in daily rainfall prediction from light gradient boosting machine (LightGBM) algorithm. Here, empirical mode decomposition (EMD) and singular spectrum analysis (SSA) methods were considered to decompose and reconstruct input time series into trend terms, fluctuating terms, and noise components. The decomposed time series from EMD and SSA methods were used as input data for LightGBM algorithm in two hybrid models, including empirical mode-based light gradient boosting machine (EMDGBM) and singular spectrum analysis-based light gradient boosting machine (SSAGBM), respectively. A total of four parameters (i.e., temperature, humidity, wind speed, and rainfall) at a daily scale from 2003 to 2017 is used as input data for daily rainfall prediction. As results from statistical performance indicators, it indicates that the SSAGBM model shows a better performance than the EMDGBM model and the original LightGBM algorithm with no decomposition methods. It represents that the accuracy of LightGBM algorithm in rainfall prediction was improved with the SSA method when using multivariate dataset.

• Geomechanics and Engineering
• /
• 제37권3호
• /
• pp.197-211
• /
• 2024

This research studies the effect of geotechnical factors on EPB-TBM performance parameters. The modeling was performed using simple and multivariate linear regression methods, artificial neural networks (ANNs), and Sugeno fuzzy logic (SFL) algorithm. In ANN, 80% of the data were randomly allocated to training and 20% to network testing. Meanwhile, in the SFL algorithm, 75% of the data were used for training and 25% for testing. The coefficient of determination (R²) obtained between the observed and estimated values in this model for the thrust force and cutterhead torque was 0.19 and 0.52, respectively. The results showed that the SFL outperformed the other models in predicting the target parameters. In this method, the R² obtained between observed and predicted values for thrust force and cutterhead torque is 0.73 and 0.63, respectively. The sensitivity analysis results show that the internal friction angle (φ) and standard penetration number (SPT) have the greatest impact on thrust force. Also, earth pressure and overburden thickness have the highest effect on cutterhead torque.

• 대한원격탐사학회지
• /
• 제32권4호
• /
• pp.383-401
• /
• 2016

In this paper, we described the statistical modeling of crop yields using satellite images, climatic datasets, soil property maps, and fertilizer data for the Midwestern United States during 2001-2012. Satellite images were obtained from the Moderate Resolution Imaging Spectroradiometer (MODIS), and climatic datasets were provided by the Parameter-elevation Regressions on Independent Slopes Model (PRISM) Climate Group. Soil property maps were derived from the Harmonized World Soil Database (HWSD). Our multivariate regression models produced quite good prediction accuracies, with differences of approximately 8-15% from the governmental statistics of corn and soybean yields. The unfavorable conditions of climate and vegetation in 2012 could have resulted in a decrease in yields according to the regression models, but the actual yields were greater than predicted. It can be interpreted that factors other than climate, vegetation, soil, and fertilizer may be involved in the negative biases. Also, we found that soybean yield was more affected by minimum temperature conditions while corn yield was more associated with photosynthetic activities. These two crops can have different potential impacts regarding climate change, and it is important to quantify the degree of the crop sensitivities to climatic variations to help adaptation by humans. Considering the yield decreases during the drought event, we can assume that climatic effect may be stronger than human adaptive capacity. Thus, further studies are demanded particularly by enhancing the data regarding human activities such as tillage, fertilization, irrigation, and comprehensive agricultural technologies.

• 한국환경보건학회지
• /
• 제42권4호
• /
• pp.280-292
• /
• 2016

Objectives: The objective of this study was to build COD regression models for the Han River and evaluate water quality. Methods: Water quality data sets for the dry season (as of January) during a four-year period (2012-2015) were collected from the database of the Han River automatic water quality monitoring stations. Statistical techniques, including combined genetic algorithm-multiple linear regression (GA-MLR) were used to build five-descriptor COD models. Multivariate statistical techniques such as principal component analysis (PCA) and cluster analysis (CA) are useful tools for extracting meaningful information. Results: The $r^2$ of the best COD models provided significant high values (> 0.8) between 2012 and 2015. Total organic carbon (TOC) was a surrogate indicator for COD (as COD/TOC) with high reliability ($r^2=0.63$ in 2012, $r^2=0.75$ for 2013, $r^2=0.79$ for 2014 and $r^2=0.85$ for 2015). The ratios of COD/TOC were calculated as 2.08 in 2012, 1.79 in 2013, 1.52 and 1.45 in 2015, indicating that biodegradability in the water body of the Han River was being sustained, thereby further improving water quality. The BOD/COD ratio supported these findings. The cluster analysis revealed higher annual levels of microorganisms and phosphorous at stations along the Hangang-Seoul and Hantangang areas. Nevertheless, the overall water quality over the last four years showed an observable trend toward continuous improvement. These findings also suggest that non-point pollution control strategies should consider the influence of upstreams and downstreams to protect water quality in the Han River. Conclusion: This data analysis procedure provided an efficient and comprehensive tool to interpret complex water quality data matrices. Results from a trend analysis provided much important information about sources and parameters for Han River water quality management.

• 응용통계연구
• /
• 제33권3호
• /
• pp.281-296
• /
• 2020

같은 개체로부터 반복 측정한 자료를 경시적 자료(longitudinal data)라고 한다. 이러한 자료를 분석하려면 흔히 사용되는 횡단 자료 분석과는 다른 분석 방법이 필요하다. 즉, 경시적 자료에서 공변량의 효과를 추정할 때에는 반복 측정된 결과 간의 상관성을 고려해야 하며, 따라서 공분산행렬을 모형화 하는 것이 매우 중요하다. 그러나 추정해야 할 모수가 많고, 추정된 공분산행렬이 양정치성을 만족해야 하므로 공분산 행렬의 모형화는 쉽지 않다. 특히 다변량 경시적 자료분석을 위한 공분산행렬의 모형화는 더욱더 심층적인 방법론을 사용해야 한다. 본 논문은 다변량 경시적 자료분석을 위한 공분산행렬을 모형화하기 위해 두 가지 방법론을 고찰한다. 두 방법 모두 수정된 콜레스키 분해(modified Cholesky decomposition)를 이용하여 시간에 따른 응답변수들의 상관관계를 설명하고 있다. 하지만 같은 시간에서 관측된 응답변수들간의 상관관계를 설명하는 방법이 다르다. 첫 번째 방법론에서는 향상된 선형 공분산 모형(enhanced linear covariance models)을 사용하여 공분산행렬이 양정치성을 만족하도록 한다. 두 번째 방법론에서는 분산-공분산 분해(variance-correlation decomposition)와 초구분해(hypersphere decomposition)을 이용하여 공분산 행렬을 모형화 한다. 이 두 방법론의 성능을 비교하고자 모의실험을 진행한다.

• Communications for Statistical Applications and Methods
• /
• 제25권2호
• /
• pp.217-234
• /
• 2018

Copulas are a tool for constructing multivariate distributions and formalizing the dependence structure between random variables. From copula literature review, there are a few asymmetric copulas available so far while data collected from the real world often exhibit asymmetric nature. This necessitates developing asymmetric copulas. In this study, we discuss a method to construct a new class of bivariate asymmetric copulas based on products of symmetric (sometimes asymmetric) copulas with powered arguments in order to determine if the proposed construction can offer an added value for modeling asymmetric bivariate data. With these newly constructed copulas, we investigate dependence properties and measure of association between random variables. In addition, the test of symmetry of data and the estimation of hyper-parameters by the maximum likelihood method are discussed. With two real example such as car rental data and economic indicators data, we perform the goodness-of-fit test of our proposed asymmetric copulas. For these data, some of the proposed models turned out to be successful whereas the existing copulas were mostly unsuccessful. The method of presented here can be useful in fields such as finance, climate and social science.

• 한국자기공명학회논문지
• /
• 제25권3호
• /
• pp.33-38
• /
• 2021

Type 1 diabetes mellitus (T1DM) is caused by insufficient production of insulin, which is involved in carbohydrate metabolism. Type 2 diabetes mellitus (T2DM) has insulin resistance in which cells do not respond adequately to insulin. The purpose of this study was to estimate the characteristics of type 1 diabetes using streptozotocin-treated mice (STZ-mouse). The sera samples were collected from the models of hyperglycemic mouse and healthy mouse. Based on the pair-wise comparison, five metabolites were found to be noticeable: glucose, malonic acid, 3-hyroxybutyrate, methanol, and tryptophan. It was very natural glucose was upregulated in STZ-mouse. 3-hyroxybutyrate was also increased in the model. However, malonic acid, tryptophan, and methanol was downregulated in STZ-mouse. Several metabolites acetoacetate, acetone, alanine, arginine, asparagine, histidine, lysine, malate, methionine, ornithine, proline, propylene glycol, threonine, tyrosine, and urea tended to be varied in STZ-mouse while the statistical significance was not stratified for the variation. The multivariate model of PCA clearly showed the group separation between healthy control and STZ-mouse. The most significant metabolites that contributed the group separation included glucose, citrate, ascorbate, and lactate. Lactate did not show the statistical significance of change in t-test while it tends to down-regulated both in DNP and Diabetes.

• 대한산업공학회지
• /
• 제33권4호
• /
• pp.393-401
• /
• 2007

Affective aspects of design attributes such as color, Pattern, and texture are important to the overall impression and the success of interior products. Among all the interior materials, wallpapers and flooring materials take up largest construction area and they are main components in creating affective impression for customers. This study aims to investigate the relationship between luxuriousness and related affective variables and design elements of wallpapers and flooring materials. The approach consists of 3 steps: (1) selecting related affective features and product design attributes through a literature survey, opinion of expert panel, and focus group interview, (2) conducting evaluation experiments, and (3) developing Kansei models using multivariate statistical analysis and analyzing critical attributes. Evaluation experiment was conducted using a questionnaire made up of 7-point scale and 100-point scale and 30 housewives and 20 interior designers participated in the evaluation experiment. The result of evaluation was analyzed through principal component regression and quantification I analysis. As a result of analyzing the survey data, the relationship between luxuriousness and related affective features and product design attributes was identified, moreover a optimal combination of the design component was identified. Consequently, it is expected that the results of the study would be a basis of the concept of emotion-based design by giving insights about how customers perceive the luxuriousness and suggesting the optimal combination, and providing specific quantitative design guidelines.