• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.228-228
• /
• 2023

국지적 지역에 내리는 강한 강도의 강우는 많은 인명 및 재산 피해를 발생시킨다. 이러한 피해를 예방하기 위해 도시 침수 예측에 관한 연구가 오랜 기간 수행되어 왔으며, 최근에는 다양한 신경망(neural network) 모델이 활발히 이용되고 있다. 강우 지속 기간이나 강도는 일정하지 않고, 공간적 특징 또한 도시마다 다르므로 안정적인 침수 예측을 위한 신경망 모델은 강건성(robustness)을 지녀야 한다. 강건한 신경망 모델이란 적대적 공격(adversarial attack)을 방어할 수 있는 능력을 갖춘 모델을 일컫는다. 따라서 본 연구에서는, 도시 침수 예측을 위한 시공간 신경망(spatio-temporal neural network) 모델의 강건성 제고를 위한 방법론을 제안한다. 먼저 적대적 공격의 유형과 방어 방법을 분류하고, 시공간 신경망 모델의 학습 데이터 특성 및 모델 구조구성 조건 등을 활용하여 최적의 강건성 제고 방안을 도출하였다. 해당 모델은 집중호우로 인해 나타날 다양한 관망에서의 침수 피해를 각각 예측하고 피해를 예방하기 위해 활용될 수 있다.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.2 s.34
• /
• pp.21-30
• /
• 2005

In general, the robust estimation method is well known for a good statistical estimator that is insensitive to small departures from the idealized assumptions for which the estimation is optimized. While there are many existing robust estimation techniques that have been proposed in the literature, two main techniques used in computer vision are M-estimators and least-median of squares (LMS). Among these. we utilized the M-estimators since they are known to provide an optimal estimation of affine motion parameters. The M-estimators have higher statistical efficiency but tolerate much lower percentages of outliers unless properly initialized. To resolve these problems, we proposed an adaptive M-estimators algorithm that effectively separates outliers from non-outliers and estimate affine model parameters, using a continuous sigmoid weight function. The experimental results show the superiority of our method.

• Journal of KIISE:Software and Applications
• /
• v.31 no.11
• /
• pp.1474-1482
• /
• 2004

In this paper, we propose an approach for extracting and filtering block motion vectors using an adaptive weight function. We first extract motion vectors from a sequence of images by using size-varibale block matching and then process them by adaptive robust estimation to filter out outliers (motion vectors out of concern). The proposed adaptive robust estimation defines a continuous sigmoid weight function. It then adaptively tunes the sigmoid function to its hard-limit as the residual errors between the model and input data are decreased, so that we can effectively separate non-outliers (motion vectors of concern) from outliers with the finally tuned hard-limit of the weight function. The experimental results show that the suggested approach is very effective in filtering block motion vectors.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2011.01a
• /
• pp.257-260
• /
• 2011

본 논문에서는 비디오 데이터로부터 물체의 초기 움직임 영역을 자동으로 검출하는 방법을 소개한다. 제안하는 시스템은 먼저 입력 영상을 받아들인 후 인접된 영상으로부터 일정 크기의 정방향의 블록 단위로 움직임을 나타내는 모션 벡터를 추출한다. 그리고 추출된 모션벡터를 아웃라이어를 제거하는 강건 예측 알고리즘에 적용하여 배경에 해당하는 모션벡터와 잡음 및 움직이는 물체에 해당하는 모션벡터를 구분한다. 그런 다음, 군집화 알고리즘을 적용하여 이동하는 물체를 나타내는 모션벡터를 군집화하고, 군집화된 모션벡터에 해당하는 영역의 크기가 일정 수치 값 이상일 때 움직이는 물체가 감지되었다고 판단한다. 본 논문의 실험에서는 제안된 물체의 움직임 감지 방법이 기존의 방법에 비해 성능이 보다 우수함을 보인다.

• Korean Journal of Construction Engineering and Management
• /
• v.19 no.3
• /
• pp.33-42
• /
• 2018

In the aftermath of an earthquake, seismic-damaged infrastructure systems loss estimation is the first step for the disaster response. However, lifeline systems' ability to supply service can be volatile by external factors such as disturbances of nearby facilities, and not by own physical issue. Thus, this research develops the bayesian model for probabilistic inference on common-cause and cascading failure of seismic-damaged lifeline systems. In addition, the authors present network robustness estimation metrics in the context of failure propagation. In order to quantify the functional loss and observe the effect of the mitigation plan, power and water supply system in Daegu-Gyeongbuk in South Korea is selected as case network. The simulation results show that reduction of cascading failure probability allows withstanding the external disruptions from a perspective of the robustness improvement. This research enhances the comprehensive understanding of how a single failure propagates to whole lifeline system performance and affected region after an earthquake.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.43 no.9
• /
• pp.781-786
• /
• 2015

An aeroelastic stability can be degraded due to an aeroelastic modeling error and a structural uncertainty. Therefore it is necessary to predict the aeroelastic stability boundary considering an aeroelastic modeling error and a structural uncertainty. Robust aeroelastic analysis was proposed to predict the aeroelastic stability boundary considering these error and uncertainty. In the present study, the robust aeroelastic modeling and analysis were performed by using the ${\mu}$ analysis technique and the aeroelastic model of the control fin with modal approach and MSA. The computer program for the robust aeroelastic analysis was developed and verified by comparing its results with those of conventional aeroelastic analysis methods.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.37 no.8
• /
• pp.975-981
• /
• 2013

A robust optimization is only one of the ways to minimize the effects of variances in design variables on the objective functions at the preliminary design stage. To predict the variances and to formulate the probabilistic constraints are the most important procedures for the robust optimization formulation. Though several methods such as the process capability index and the six sigma technique were proposed for the prediction and formulation of the variances and probabilistic constraints, respectively, there are few attempts using a percent defective which has been widely applied in the quality control of the manufacturing process for probabilistic constraints. In this study, the robust optimization for a lower control arm of automobile vehicle was carried out, in which the design space showing the mean and variance sensitivity of weight and stress was explored before robust optimization for a lower control arm. The 2nd order Taylor expansion for calculating the standard deviation was used to improve the numerical accuracy for predicting the variances. Simplex algorithm which does not use the gradient information in optimization was used to convert constrained optimization into unconstrained one in robust optimization.

• Journal of KIISE:Software and Applications
• /
• v.33 no.2
• /
• pp.252-266
• /
• 2006

Many problems in computer vision are mainly based on mathematical models. Their optimal solutions can be found by estimating the parameters of each model. However, provided an input data set is involved outliers which are relative]V larger than normal noises, they lead to incorrect results. RANSAC is a representative robust algorithm which is used to resolve the problem. One major problem with RANSAC is that it needs priori knowledge(i.e. a percentage of outliers) of the distribution of data. To solve this problem, we propose a FRANSAC algorithm which improves the rejection rate of outliers and the accuracy of solutions. This is peformed by categorizing all data into good sample set, bad sample set and vague sample set using a fuzzy classification at each iteration and sampling in only good sample set. In the experimental results, we show that the performance of the proposed algorithm when it is applied to the linear regression and the calculation of a homography.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.633-637
• /
• 2023

암시적 비윤리 표현은 명시적 비윤리 표현과 달리 학습 데이터 선별이 어려울 뿐만 아니라 추가 생산 패턴 예측이 까다롭다. 고로 암시적 비윤리 표현에 대한 언어 모델의 감지 능력을 기르기 위해서는 모델의 취약성을 발견하는 연구가 반드시 선행되어야 한다. 본 논문에서는 암시적 비윤리 표현에 대한 표기 변경과 긍정 요소 삽입이라는 두 가지 변형을 통해 모델의 예측 변화를 유도하였다. 그 결과 모델이 야민정음과 외계어를 사용한 언어 변형에 취약하다는 사실을 발견하였다. 이에 더해 이모티콘이 텍스트와 함께 사용되는 경우 텍스트 자체보다 이모티콘의 효과가 더 크다는 사실을 밝혀내었다.

• The Korean Journal of Applied Statistics
• /
• v.20 no.2
• /
• pp.267-280
• /
• 2007

Various single-valued design optimality criteria such as D-, G-, and V-optimality are used often in constructing optimal experimental designs for mixture experiments in a constrained region R where lower and upper bound constraints are imposed on the ingredients proportions. Even though they are optimal in the strict sense of particular optimality criterion used, it is known that their performance is unsatisfactory with respect to the prediction capability over a constrained region. (Vining et at., 1993; Khuri et at., 1999) We assume the quadratic polynomial model as the mixture response surface model and are interested in finding efficient designs in the constrained design space for a mixture. In this paper, we make an expanded list of candidate design points by adding interior points to the extreme vertices, edge midpoints, constrained face centroids and the overall centroid. Then, we want to propose a robust design with respect to D-optimality, G-optimality, V-optimality and distance-based U-optimality. Comparing scaled prediction variance quantile plots (SPVQP) of robust designs with that of recommended designs in Khuri et al. (1999) and Vining et al. (1993) in the well-known examples of a four-component fertilizer experiment as well as McLean and Anderson's Railroad Flare Experiment, robust designs turned out to be superior to those recommended designs.