• 한국지하수토양환경학회지:지하수토양환경
• /
• 제19권4호
• /
• pp.31-44
• /
• 2014

In the present study, a few of recently developed geostatistical models are comparatively studied. The models are two-point statistics based sequential indicator simulation (SISIM) and generalized coupled Markov chain (GCMC), multi-point statistics single normal equation simulation (SNESIM), and object based model of FLUVSIM (fluvial simulation) that predicts structures of target object from the provided geometric information. Out of the models, SNESIM and FLUVSIM require additional information other than conditioning data such as training map and geometry, respectively, which generally claim demanding additional resources. For the comparative studies, three-dimensional fluvial reservoir model is developed considering the genetic information and the samples, as input data for the models, are acquired by mimicking realistic sampling (i.e. random sampling). For SNESIM and FLUVSIM, additional training map and the geometry data are synthesized based on the same information used for the objective model. For the comparisons of the predictabilities of the models, two different measures are employed. In the first measure, the ensemble probability maps of the models are developed from multiple realizations, which are compared in depth to the objective model. In the second measure, the developed realizations are converted to hydrogeologic properties and the groundwater flow simulation results are compared to that of the objective model. From the comparisons, it is found that the predictability of GCMC outperforms the other models in terms of the first measure. On the other hand, in terms of the second measure, the both predictabilities of GCMC and SNESIM are outstanding out of the considered models. The excellences of GCMC model in the comparisons may attribute to the incorporations of directional non-stationarity and the non-linear prediction structure. From the results, it is concluded that the various geostatistical models need to be comprehensively considered and comparatively analyzed for appropriate characterizations.

• 한국가스학회지
• /
• 제24권6호
• /
• pp.91-97
• /
• 2020

리튬이온 배터리(LIB)는 다른 배터리에 비해 수명이 길고, 에너지 밀도가 높으며, 자체 방전율이 낮아, 에너지 저장장치(ESS)로 선호되고 있다. 하지만, 2017~2019년 기간 동안 국내에서만도 28건의 화재사고가 발생하였으며, LIB의 운영 중 안전성 및 신뢰성을 보장하기 위해 LIB의 정확한 용량추정은 필수요소이다. 본 연구에서는 LIB의 충방전 cycle에 따른 용량변화를 예측하는 기계학습 기반 모델의 설계에 있어 중요한 요소인 최적 머신러닝 모델의 선정을 위해, Decision Tree, 앙상블학습법, Support Vector Regression, Gaussian Process Regression (GPR) 각각을 이용한 예측모델을 구현하고 성능비교를 실시하였다. 학습을 위해 NASA에서 제공하는 시험데이터를 사용하였으며, GPR이 가장 좋은 예측성능을 보였다. 이를 바탕으로 추가 시험데이터 학습을 통해 개선된 LIB 용량예측과 잔여 수명추정 모델을 개발하여, 운영 중 이상 감지 및 모니터링 성능을 높여, 보다 안전하고 안정된 ESS 운용에 활용하고자 한다.

• 스마트미디어저널
• /
• 제12권2호
• /
• pp.9-14
• /
• 2023

이 논문에서는 gradCAM를 활용한 적은 데이터로 얼굴 전체 또는 더 다양한 feature을 사용하여 얼굴인식을 할 수 있는 새로운 앙상블 방법론을 제안하였다. 인공지능 모델의 판단 근거는 gradCAM을 통하여 saliency map으로 표현될 수 있다. 따라서 본 논문에서는 학습된 얼굴인식 모델이 어느 부분에 편향적으로 관찰하여 판단했는지 gradCAM으로 시각화한다. 계산된 saliency map에서 일정 수치 이상의 돌출된 부분을 추가 모델이 학습에 사용할 수 없도록 노이즈를 추가해 데이터를 생산한다. 노이즈를 추가해서 만든 데이터로 학습할 경우 노이즈 부분을 활용하여 학습을 할 수 없으므로 새로운 얼굴 부분을 사용하여 얼굴인식 네트워크를 학습하게 된다. 기본 데이터로 학습한 네트워크와 돌출 부분에 노이즈를 추가해서 학습한 모델은 얼굴의 서로 다른 얼굴 feature을 사용할 수밖에 없고, 앙상블로 결합했을 때 얼굴의 좀 더 다양한 부분들을 사용한 임베딩 feature를 만들 수 있다. 이 논문에서 제안하는 앙상블 기법은 일반적인 앙상블 모델보다 정확도는 1.79% 상승하였고 equal error rate (EER)은 0.01788 감소하였다.

• 대한산업공학회지
• /
• 제28권2호
• /
• pp.147-161
• /
• 2002

In novelty detection, one attempts to discriminate abnormal patterns from normal ones. Novelty detection is quite difficult since, unlike usual two class classification problems, only normal patterns are available for training. Auto-Associative Multi-Layer Perceptron (AAMLP) has been shown to provide a good performance based upon the property that novel patterns usually have larger auto-associative errors. In this paper, we give a mathematical analysis of 2-layer AAMLP's output characteristics and empirical results of 2-layer and 4-layer AAMLPs. Various activation functions such as linear, saturated linear and sigmoid are compared. The 2-layer AAMLPs cannot identify non-linear boundaries while the 4-layer ones can. When the data distribution is multi-modal, then an ensemble of AAMLPs, each of which is trained with pre-clustered data is required. This paper contributes to understanding of AAMLP networks and leads to practical recommendations regarding its use.

• 한국융합학회논문지
• /
• 제9권12호
• /
• pp.9-15
• /
• 2018

확장된 데이터 표현의 주요 목표는 유비쿼터스 환경에서 일반적인 문제에 적합한 데이터 구조를 개발하는 것이다. 이 방법의 가장 큰 특징은 속성 값을 확률로 표현할 수 있다는 것이다. 다음 특성은 훈련 데이터의 각 이벤트가 중요도를 나타내는 가중치 값을 갖도록 한다는 것이다. 데이터 구조가 개발된 후에 이를 학습할 수 있는 알고리즘이 고안된다. 그 동안 이 알고리즘은 여러 분야에서 여러 문제에 적용하여 좋은 결과를 산출해 왔다. 본 논문은 먼저 데이터 표현 기법인 UChoo를 소개하고 이론적인 배경이 되는 규칙 개선 문제를 소개한다. 그리고 규칙 개선, 손실 데이터 처리, BEWS 문제, 앙상블 시스템과 같은 응용 분야의 예를 소개한다.

• 한국운동역학회지
• /
• 제33권3호
• /
• pp.85-93
• /
• 2023

Objective: This study examined differences in joint kinematics and movement variability of lower extremity between adolescent athletes with and without lateral ankle sprain (LAS) history during drop vertical jump. Method: Fourteen adolescent athletes with LAS history and 14 controls participated in this study. The independent variable was group while dependent variables were 3D joint kinematics and movement variability of hip, knee, and ankle joint. Ensemble curve analyses were conducted to identify differences in movement strategies between two groups. Results: The LAS group showed that greater eversion during jump phase compared with the control group. Additionally, less movement variability was found in the LAS group during the pre-landing and jump phases in ankle and hip joints compared with the control group. Conclusion: The LAS group may adapt the environmental constraints by reducing the movement variability in ankle and hip joints. However, training programs focusing on recovery of ankle function should be emphasized after LAS because excessive pronation for prevention of LAS during the jump phase may result in reduced performance.

• 한국공간구조학회논문집
• /
• 제23권3호
• /
• pp.87-94
• /
• 2023

Chloride is one of the most common threats to reinforced concrete (RC) durability. Alkaline environment of concrete makes a passive layer on the surface of reinforcement bars that prevents the bar from corrosion. However, when the chloride concentration amount at the reinforcement bar reaches a certain level, deterioration of the passive protection layer occurs, causing corrosion and ultimately reducing the structure's safety and durability. Therefore, understanding the chloride diffusion and its prediction are important to evaluate the safety and durability of RC structure. In this study, the chloride diffusion coefficient is predicted by machine learning techniques. Various machine learning techniques such as multiple linear regression, decision tree, random forest, support vector machine, artificial neural networks, extreme gradient boosting annd k-nearest neighbor were used and accuracy of there models were compared. In order to evaluate the accuracy, root mean square error (RMSE), mean square error (MSE), mean absolute error (MAE) and coefficient of determination (R2) were used as prediction performance indices. The k-fold cross-validation procedure was used to estimate the performance of machine learning models when making predictions on data not used during training. Grid search was applied to hyperparameter optimization. It has been shown from numerical simulation that ensemble learning methods such as random forest and extreme gradient boosting successfully predicted the chloride diffusion coefficient and artificial neural networks also provided accurate result.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.728-735
• /
• 2022

Roofers get exposed to increased risk of knee musculoskeletal disorders (MSDs) at different phases of a sloped shingle installation task. As different phases are associated with different risk levels, this study explored the application of machine learning for automated classification of seven phases in a shingle installation task using knee kinematics and roof slope information. An optical motion capture system was used to collect knee kinematics data from nine subjects who mimicked shingle installation on a slope-adjustable wooden platform. Four features were used in building a phase classification model. They were three knee joint rotation angles (i.e., flexion, abduction-adduction, and internal-external rotation) of the subjects, and the roof slope at which they operated. Three ensemble machine learning algorithms (i.e., random forests, decision trees, and k-nearest neighbors) were used for training and prediction. The simulations indicate that the k-nearest neighbor classifier provided the best performance, with an overall accuracy of 92.62%, demonstrating the considerable potential of machine learning methods in detecting shingle installation phases from workers knee joint rotation and roof slope information. This knowledge, with further investigation, may facilitate knee MSD risk identification among roofers and intervention development.

• Computers and Concrete
• /
• 제33권2호
• /
• pp.137-145
• /
• 2024

The technique of experimentally determining concrete's compressive strength for a given mix design is time-consuming and difficult. The goal of the current work is to propose a best working predictive model based on different machine learning algorithms such as Gradient Boosting Machine (GBM), Stacked Ensemble (SE), Distributed Random Forest (DRF), Extremely Randomized Trees (XRT), Generalized Linear Model (GLM), and Deep Learning (DL) that can forecast the compressive strength of ternary geopolymer concrete mix without carrying out any experimental procedure. A geopolymer mix uses supplementary cementitious materials obtained as industrial by-products instead of cement. The input variables used for assessing the best machine learning algorithm not only include individual ingredient quantities, but molarity of the alkali activator and age of testing as well. Myriad statistical parameters used to measure the effectiveness of the models in forecasting the compressive strength of ternary geopolymer concrete mix, it has been found that GBM performs better than all other algorithms. A sensitivity analysis carried out towards the end of the study suggests that GBM model predicts results close to the experimental conditions with an accuracy between 95.6 % to 98.2 % for testing and training datasets.

• 한국음향학회지
• /
• 제40권5호
• /
• pp.479-487
• /
• 2021

본 논문에서는 잡음 학생 모델 기반의 자가 학습을 활용한 음향 사건 검지 기법을 제안한다. 제안된 음향 사건 검지 모델은 두 단계로 구성된다. 첫 번째 단계에서는 잔차 합성곱 순환 신경망(Residual Convolutional Recurrent Neural Network, RCRNN)을 훈련하여 레이블이 지정되지 않은 비표기 데이터셋의 레이블 예측에 활용한다. 두 번째 단계에서는 세 가지 잡음 종류를 적용한 잡음 학생 모델을 자가학습 기법으로 반복하여 학습한다. 여기서 잡음 학생 모델은 SpecAugment, Mixup, 시간-주파수 이동을 활용한 특징 잡음, 드롭아웃을 활용한 모델 잡음, 그리고 semi-supervised loss function을 적용한 레이블 잡음을 활용하여 학습된다. 제안된 음향 사건 검지 모델의 성능은 Detection and Classification of Acoustic Scenes and Events(DCASE) 2020 Challenge Task 4의 validation set으로 평가하였다. DCASE 2020 챌린지 데이터셋의 baseline 및 최상위 랭크된 모델과 이벤트 단위 F1 점수 성능을 비교한 결과, 제안된 음향 사건 검지 모델이 단일 모델과 앙상블 모델에서 최상위 모델 대비 F1 점수를 각각 4.6 %와 3.4 % 향상시켰다.