• Steel and Composite Structures
• /
• 제47권6호
• /
• pp.759-779
• /
• 2023

The paper proposes two hybrid metaheuristic optimization and artificial neural network (ANN) methods for the close prediction of the ultimate axial compressive capacity of concentrically loaded concrete filled double skin steel tube (CFDST) columns. Two metaheuristic optimization, namely genetic algorithm (GA) and particle swarm optimization (PSO), approaches enable the dynamic training architecture underlying an ANN model by optimizing the number and sizes of hidden layers as well as the weights and biases of the neurons, simultaneously. The former is termed as GA-ANN, and the latter as PSO-ANN. These techniques utilize the gradient-based optimization with Bayesian regularization that enhances the optimization process. The proposed GA-ANN and PSO-ANN methods construct the predictive ANNs from 125 available experimental datasets and present the superior performance over standard ANNs. Both the hybrid GA-ANN and PSO-ANN methods are encoded within a user-friendly graphical interface that can reliably map out the accurate ultimate axial compressive capacity of CFDST columns with various geometry and material parameters.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2023년도 학술발표회
• /
• pp.29-29
• /
• 2023

To mitigate the damaging impacts of floods, accurate prediction of runoff, streamflow and flood inundation is needed. Conventional approach of simulating hydrology and hydraulics using loosely coupled models cannot capture the complex dynamics of surface and sub-surface processes. Additionally, the scarcity of data in ungauged basins and quality of data in gauged basins add uncertainty to model predictions, which need to be quantified. In this presentation, first the role of integrated modeling on creating accurate flood simulations and inundation maps will be presented with specific focus on urban environments. Next, the use of machine learning in producing streamflow predictions will be presented with specific focus on incorporating covariate shift and the application of theory guided machine learning. Finally, a framework to quantify the uncertainty in flood models using Hierarchical Bayesian Modeling Averaging will be presented. Overall, this presentation will highlight that creating accurate information on flood magnitude and extent requires innovation and advancement in different aspects related to hydrologic predictions.

• Wind and Structures
• /
• 제36권6호
• /
• pp.423-434
• /
• 2023

Aerodynamic force coefficients are generally obtained from traditional wind tunnel tests or computational fluid dynamics (CFD). Unfortunately, the techniques mentioned above can sometimes be cumbersome because of the cost involved, such as the computational cost and the use of heavy equipment, to name only two examples. This study proposed to build a deep neural network model to predict the aerodynamic force coefficients based on data collected from CFD simulations to overcome these drawbacks. Therefore, a series of CFD simulations were conducted using different geometric parameters to obtain the aerodynamic force coefficients, validated with wind tunnel tests. The results obtained from CFD simulations were used to create a dataset to train a multilayer perceptron artificial neural network (ANN) model. The models were obtained using three optimization algorithms: scaled conjugate gradient (SCG), Bayesian regularization (BR), and Levenberg-Marquardt algorithms (LM). Furthermore, the performance of each neural network was verified using two performance metrics, including the mean square error and the R-squared coefficient of determination. Finally, the ANN model proved to be highly accurate in predicting the force coefficients of similar bridge sections, thus circumventing the computational burden associated with CFD simulation and the cost of traditional wind tunnel tests.

• 한국통신학회논문지
• /
• 제36권4B호
• /
• pp.354-361
• /
• 2011

스마트 홈 분야에 대한 관심의 증가와 기술적인 발전은 상황인지 서비스와 베이시안 네트워크 알고리즘, 트리 구조 알고리즘 그리고 유전자 추측 알고리즘과 같은 예측 알고리즘에 대한 연구가 활발히 이루어지고 있다. 상황 인지 서비스는 개별적인 사용자의 패턴을 고려한 맞춤형 서비스를 제공하는 것은 사용자의 삶의 질을 향상시키는 데 도움 주는 것을 의미한다. 하지만 상황인지 서비스를 구현 하는 것은 상황정보와의 부합성 문제와 예외적인 상황 처리가 고려하는데 어려움을 겪고 있다. 이 문제를 해결하기 위해서, 본 연구에서는 지능형 순차적 매칭 방식 알고리즘(Intelligent Sequential Matching Algorithm : ISMA)을 제안하고, 페트리 넷에 시간 개념을 추가하여 시간 페트리 넷(Timed Petri-net : TPN)으로 모델링한다. 제안한 지능형 순차적 매칭 알고리즘의 유효성을 증명하기 위하여 시나리오를 제시하고, 그것을 모델링 한다. 또한 동일한 실험 조건 아래, 기존의 예측 알고리즘과 비교를 통하여 과 제안된 알고리즘의 예측 정확도가 4~6% 우수함을 보인다.

• 응용통계연구
• /
• 제30권6호
• /
• pp.917-929
• /
• 2017

한국노동패널조사에서 제공하는 2015년 한국 생산가능인구의 월평균 소득분포를 보면 0 관측치의 비율이 과도하게 높은 형태를 보여 기존의 소득분포에 주로 사용되는 토빗모형으로는 설명에 한계가 있다. 본 연구에서는 영과잉 특성을 반영하여 영과잉 토빗모형을 사용하여 한국인의 소득 자료를 분석한다. 영과잉 토빗모형은 2단계 모형으로 1단계에서는 소득이 0인 그룹을 두 그룹으로 나누는데, 첫 번째 그룹은 노동시장 참여의지가 없어 시장에 참여하지 않으므로 0이 관측되는 그룹(genuine zero)이고 두 번째 그룹은 노동시장 참여의지는 있으나 낮은 임금으로 인하여 절단되어 0이 관측되는 그룹(random zero)으로 가정하였다. 두 번째 random zero 그룹은 0 이상의 연속 자료와 결합하여 토빗모형을 적용한다. 1단계와 2단계 모형에 관심 있는 설명변수를 가진 회귀모형을 적용하여 노동시장 참여여부와 임금 수준에 영향을 미치는 요인을 알아본다. 마코브 체인 몬테칼로 기법을 사용하여 모수를 추정하고 기존의 토빗모형과 비교한 결과 영과잉 토빗모형이 0의 빈도추정과 모형 적합도 면에서 우수한 결과를 보였다. 분석결과 나이가 많을수록, 남자가 여자보다, 학력이 낮을수록, 노동시장에 참여할 가능성이 매우 유의하게 높으며, 사회경제적 지위가 높을수록 그리고 유보임금이 낮을수록 노동시장에 참여하지 않을 확률이 높은 것으로 나타났다. 임금수준을 보면, 남자가 여자보다, 학력이 높을수록, 기혼이 미혼 보다 매우 유의하게 더 높은 임금을 받는 것으로 나타났다.

• 인터넷정보학회논문지
• /
• 제22권2호
• /
• pp.77-87
• /
• 2021

인터넷과 개인용 컴퓨터가 발달하면서 다양하고 복잡한 공격들이 등장하기 시작했다. 공격들이 복잡해짐에 따라 기존에 사용하던 시그니처 기반의 탐지 방식으로 탐지가 어려워졌으며 이를 해결하기 위해 행위기반의 탐지를 위한 로그 이상탐지에 대한 연구가 주목 받기 시작했다. 최근 로그 이상탐지에 대한 연구는 딥러닝을 활용해 순서를 학습하는 방식으로 이루어지고 있으며 좋은 성능을 보여준다. 하지만 좋은 성능에도 불구하고 판단에 대한 근거를 제공하지 못한다는 한계점을 지닌다. 판단에 대한 근거 및 설명을 제공하지 못할 경우, 데이터가 오염되거나 모델 자체에 결함이 발생해도 이를 발견하기 어렵다는 문제점을 지닌다. 결론적으로 사용자의 신뢰성을 잃게 된다. 이를 해결하기 위해 본 연구에서는 설명가능한 로그 이상탐지 시스템을 제안한다. 본 연구는 가장 먼저 로그 파싱을 진행해 로그 전처리를 수행한다. 이후 전처리된 로그들을 이용해 베이지안 확률 기반 순차 규칙추출을 진행한다. 결과적으로 "If 조건 then 결과, 사후확률(θ)" 형식의 규칙집합을 추출하며 이와 매칭될 경우 정상, 매칭되지 않을 경우, 이상행위로 판단하게 된다. 실험으로는 HDFS 로그 데이터셋을 활용했으며, 그 결과 F1score 92.7%의 성능을 나타내었다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제13권3호
• /
• pp.186-199
• /
• 2013

We pose pattern classification as a density estimation problem where we consider mixtures of generative models under partially labeled data setups. Unlike traditional approaches that estimate density everywhere in data space, we focus on the density along the decision boundary that can yield more discriminative models with superior classification performance. We extend our earlier work on the recursive estimation method for discriminative mixture models to semi-supervised learning setups where some of the data points lack class labels. Our model exploits the mixture structure in the functional gradient framework: it searches for the base mixture component model in a greedy fashion, maximizing the conditional class likelihoods for the labeled data and at the same time minimizing the uncertainty of class label prediction for unlabeled data points. The objective can be effectively imposed as individual mixture component learning on weighted data, hence our mixture learning typically becomes highly efficient for popular base generative models like Gaussians or hidden Markov models. Moreover, apart from the expectation-maximization algorithm, the proposed recursive estimation has several advantages including the lack of need for a pre-determined mixture order and robustness to the choice of initial parameters. We demonstrate the benefits of the proposed approach on a comprehensive set of evaluations consisting of diverse time-series classification problems in semi-supervised scenarios.

• International Journal of Naval Architecture and Ocean Engineering
• /
• 제11권1호
• /
• pp.22-32
• /
• 2019

On offshore platforms, oil and gas leaks are apt to be the initial events of major accidents that may result in significant loss of life and property damage. To prevent accidents induced by leakage, it is vital to perform a case-specific and accurate risk assessment. This paper presents an integrated method of Ddynamic Qquantitative Rrisk Aassessment (DQRA)-using the Decision Making Trial and Evaluation Laboratory (DEMATEL)-Bayesian Network (BN)-for evaluation of the system vulnerabilities and prediction of the occurrence probabilities of accidents induced by leakage. In the method, three-level indicators are established to identify factors, events, and subsystems that may lead to leakage, fire, and explosion. The critical indicators that directly influence the evolution of risk are identified using DEMATEL. Then, a sequential model is developed to describe the escalation of initial events using an Event Tree (ET), which is converted into a BN to calculate the posterior probabilities of indicators. Using the newly introduced accident precursor data, the failure probabilities of safety barriers and basic factors, and the occurrence probabilities of different consequences can be updated using the BN. The proposed method overcomes the limitations of traditional methods that cannot effectively utilize the operational data of platforms. This work shows trends of accident risks over time and provides useful information for risk control of floating marine platforms.

• 반도체디스플레이기술학회지
• /
• 제22권1호
• /
• pp.134-138
• /
• 2023

This study focuses on the analysis of the results of computational fluid dynamics simulations of mist-chemical vapor deposition for the growth of an epitaxial wafer in power semiconductor technology using artificial intelligence techniques. The conventional approach of predicting the uniformity of the deposited layer using computational fluid dynamics and design of experimental takes considerable time. To overcome this, artificial intelligence method, which is widely used for optimization, automation, and prediction in various fields, was utilized to analyze the computational fluid dynamics simulation results. The computational fluid dynamics simulation results were analyzed using a supervised deep neural network model for regression analysis. The predicted results were evaluated quantitatively using Euclidean distance calculations. And the Bayesian optimization was used to derive the optimal condition, which results obtained through deep neural network training showed a discrepancy of approximately 4% when compared to the results obtained through computational fluid dynamics analysis. resulted in an increase of 146.2% compared to the previous computational fluid dynamics simulation results. These results are expected to have practical applications in various fields.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2015년도 학술발표회
• /
• pp.225-225
• /
• 2015

Drought events usually evolve slowly in time and their impacts generally span a long period of time. This indicates that the sequence of drought is not completely random. The Hidden Markov Model (HMM) is a probabilistic model used to represent dependences between invisible hidden states which finally result in observations. Drought characteristics are dependent on the underlying generating mechanism, which can be well modelled by the HMM. This study employed a HMM with Gaussian emissions to fit the Standardized Precipitation Index (SPI) series and make multi-step prediction to check the drought characteristics in the future. To estimate the parameters of the HMM, we employed a Bayesian model computed via Markov Chain Monte Carlo (MCMC). Since the true number of hidden states is unknown, we fit the model with varying number of hidden states and used reversible jump to allow for transdimensional moves between models with different numbers of states. We applied the HMM to several stations SPI data in South Korea. The monthly SPI data from January 1973 to December 2012 was divided into two parts, the first 30-year SPI data (January 1973 to December 2002) was used for model calibration and the last 10-year SPI data (January 2003 to December 2012) for model validation. All the SPI data was preprocessed through the wavelet denoising and applied as the visible output in the HMM. Different lead time (T= 1, 3, 6, 12 months) forecasting performances were compared with conventional forecasting techniques (e.g., ANN and ARMA). Based on statistical evaluation performance, the HMM exhibited significant preferable results compared to conventional models with much larger forecasting skill score (about 0.3-0.6) and lower Root Mean Square Error (RMSE) values (about 0.5-0.9).