• International Journal of Fluid Machinery and Systems
• /
• 제10권3호
• /
• pp.240-253
• /
• 2017

Latin hypercube sampling is widely used design-of-experiment technique to select design points for simulation which are then used to construct a surrogate model. The exploration/exploitation properties of surrogate models depend on the size and distribution of design points in the chosen design space. The present study aimed at evaluating the performance characteristics of various surrogate models depending on the Latin hypercube sampling (LHS) procedure (sample size and spatial distribution) for a diverse set of optimization problems. The analysis was carried out for two types of problems: (1) thermal-fluid design problems (optimizations of convergent-divergent micromixer coupled with pulsatile flow and boot-shaped ribs), and (2) analytical test functions (six-hump camel back, Branin-Hoo, Hartman 3, and Hartman 6 functions). The three surrogate models, namely, response surface approximation, Kriging, and radial basis neural networks were tested. The important findings are illustrated using Box-plots. The surrogate models were analyzed in terms of global exploration (accuracy over the domain space) and local exploitation (ease of finding the global optimum point). Radial basis neural networks showed the best overall performance in global exploration characteristics as well as tendency to find the approximate optimal solution for the majority of tested problems. To build a surrogate model, it is recommended to use an initial sample size equal to 15 times the number of design variables. The study will provide useful guidelines on the effect of initial sample size and distribution on surrogate construction and subsequent optimization using LHS sampling plan.

• 대한의용생체공학회:의공학회지
• /
• 제36권4호
• /
• pp.103-108
• /
• 2015

Discrimination of Parkinson's disease (PD) from Essential tremor (ET) is often misdiagnosed in clinical practice. Since tremor is time-varying signal, and dominant and harmonic frequencies are shown in tremor only with moderate or severe symptom, there are some limitations to use frequency related features. Moreover, patients with PD or ET can suffer from both resting tremor and postural tremor. In this study, 28 patients with PD and 17 patients with ET were enrolled. Tremor was measured with accelerations on the more affected hand during resting and postural conditions. The ratio of root mean square (RMS) of resting tremor to RMS of postural tremor, the mean coefficients of autocorrelation function (ACF), and the mean of differences of two adjacent coefficients of ACF at resting and postural were calculated and compared between PD and ET. The performance showed 98% accuracy with support vector machine and leave-one-out cross validation. In addition, the method accurately differentiated the patients with tremor-dominant PD from patients with ET, with 100% accuracy. Therefore, the developed algorithm can assist clinicians in diagnosing and categorizing patients with tremor, especially, patients with mild symptom or the early stage of a disease, for proper treatment.

• 대한의용생체공학회:의공학회지
• /
• 제41권5호
• /
• pp.185-194
• /
• 2020

Arrhythmia is basically diagnosed with the electrocardiogram (ECG) signal, however, ECG is difficult to measure and it requires expert help in analyzing the signal. On the other hand, the radial pulse can be measured with easy and uncomplicated way in daily life, and could be suitable bio-signal for the recent untact paradigm and extensible signal for diagnosis of Korean medicine based on pulse pattern. In this study, we developed an irregular radial pulse detection algorithm based on a learning model and considered its applicability as arrhythmia screening. A total of 1432 pulse waves including irregular pulse data were used in the experiment. Three data sets were prepared with minimal preprocessing to avoid the heuristic feature extraction. As classification algorithms, elastic net logistic regression, random forest, and extreme gradient boosting were applied to each data set and the irregular pulse detection performances were estimated using area under the receiver operating characteristic curve based on a 10-fold cross-validation. The extreme gradient boosting method showed the superior performance than others and found that the classification accuracy reached 99.7%. The results confirmed that the proposed algorithm could be used for arrhythmia screening. To make a fusion technology integrating western and Korean medicine, arrhythmia subtype classification from the perspective of Korean medicine will be needed for future research.

• 응용통계연구
• /
• 제30권6호
• /
• pp.867-876
• /
• 2017

오늘날 관찰값의 개수에 비해 변수의 개수가 큰 희박한 데이터셋은 다양한 분야에서 쉽게 찾아볼 수 있고, 통계학에서 그러한 데이터셋에 대한 분석은 하나의 도전이 되어 왔다. 그러한 희박한 데이터에 대한 분류를 위해 판별분석모형들이 최근에 개발되었다. 그러한 판별분석모형들 중 하나의 접근법은 그룹들을 잘 구분해주는 차원들을 찾기를 시도하는데, 그러한 차원들은 데이터의 변수의 개수보다 훨씬 적다. 그러한 모형에서 차원의 수는 예측과 자료의 시각화를 위해 중요한 역할을 하고 일반적으로 K-묶음 교차타당성 방법에 의해 결정된다. 하지만, 희박한 데이터의 경우 K-묶음 교차타당성 방법 적용시 각 묶음에 대한 관찰값의 개수가 매우 적을 수 있기 때문에 교차타당성에 의한 차원 수 결정은 신뢰성이 떨어질 수 있다. 따라서, 본 연구에서는 그러한 희박판별분석모형에 의해 찾아진 차원들에서 각 그룹들의 평균 간의 표준화된 거리에 근거한 측도를 사용하여 최적의 차원 수를 결정하는 방법을 제안하고, 제안된 방법은 모의실험을 통해 검증된다.

• 한국소성가공학회:학술대회논문집
• /
• 한국소성가공학회 2005년도 추계학술대회 논문집
• /
• pp.410-413
• /
• 2005

The results of an experimental and analytical study of composite pressure hull on buckling pressure are presented for LRN 300. Composite tensile test was done to know the composite material properties applied FE analysis for URN composite. We predicted the buckling and post buckling analysis of composite laminated cylindrical panels under external compression by using ABAQUS /Standard[Ver 6.4]. To obtain nonlinear static equilibrium solutions for unstable problems, where the load-displacement response can exhibit the type of nonlinear buckling behavior, during periods of the response, the load and/or the displacement may decrease as the solution evolves, used the modified Riks method. The modified Riks method is an algorithm that allows effective solution of such cases [7]. Experiments were conducted to verify the validation of present analysis for cross-ply laminated shells. The shells considered in the study have two different lamination patterns, $[{\pm}45/0/90]_{18s\;and}\;[/0/90]_{18s}$. Cylindrical panel of experiment and analysis have the radius of 200mm, length of 210mm and 60 degree of cutting angle. The critical load from experiment is $69\%$ of that of numerical analysis, because the fracture of matrix was generated before buckling. So URN 300 is not proper to use at the condition under high external pressue.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권7호
• /
• pp.2938-2956
• /
• 2020

Objective image quality assessment (IQA) models have been developed by effective features to imitate the characteristics of human visual system (HVS). Actually, HVS is extremely sensitive to color degradation and complex texture changes. In this paper, we firstly reveal that many existing full reference image quality assessment (FR-IQA) methods can hardly measure the image quality with contrast and masking texture changes. To solve this problem, considering texture masking effect, we proposed a novel FR-IQA method, called Texture and Color Quality Index (TCQI). The proposed method considers both in the masking effect texture and color visual perceptual threshold, which adopts three kinds of features to reflect masking texture, color difference and structural information. Furthermore, random forest (RF) is used to address the drawbacks of existing pooling technologies. Compared with other traditional learning-based tools (support vector regression and neural network), RF can achieve the better prediction performance. Experiments conducted on five large-scale databases demonstrate that our approach is highly consistent with subjective perception, outperforms twelve the state-of-the-art IQA models in terms of prediction accuracy and keeps a moderate computational complexity. The cross database validation also validates our approach achieves the ability to maintain high robustness.

• 한국산업정보학회논문지
• /
• 제19권2호
• /
• pp.73-83
• /
• 2014

다양한 오디오 편집 기술이 개발됨으로써 오디오 데이터의 변경이 보다 쉬워지고 그 결과로 위변조 같은 다양한 사회 문제가 발생하고 있다. 현재 이런 문제를 해결하기 위해 디지털 포렌식 기술이 활발히 연구되어지고 있다. 본 논문에서는 이러한 디지털 포렌식 기술 중의 하나로 모바일 기기를 판별하는 방법을 제안하였다. 제안 방법에서는 사람에게는 들리지 않지만 기기의 디자인과 IC로부터 발생하는 노이즈 특징을 이용한다. 위너필터를 사용하여 기기의 노이즈 음을 추출하고 MIRtoolbox를 이용하여 특징들을 추출한 후 이를 다층 신경망에 학습시켜 기기를 판별한다. 총 6개의 모바일 기기를 사용하였으며 5-fold test를 통하여 99.9%의 판별 성능을 보였다. 또한 UCC 사이트에 업로드 된 데이터에서도 노이즈 음을 통한 판별이 가능한지 실험을 진행하였으며 99.8%의 판별 성능을 보였다.

• 천문학회지
• /
• 제45권2호
• /
• pp.31-38
• /
• 2012

In this study we apply Support Vector Machine (SVM) to the prediction of geo-effective halo coronal mass ejections (CMEs). The SVM, which is one of machine learning algorithms, is used for the purpose of classification and regression analysis. We use halo and partial halo CMEs from January 1996 to April 2010 in the SOHO/LASCO CME Catalog for training and prediction. And we also use their associated X-ray flare classes to identify front-side halo CMEs (stronger than B1 class), and the Dst index to determine geo-effective halo CMEs (stronger than -50 nT). The combinations of the speed and the angular width of CMEs, and their associated X-ray classes are used for input features of the SVM. We make an attempt to find the best model by using cross-validation which is processed by changing kernel functions of the SVM and their parameters. As a result we obtain statistical parameters for the best model by using the speed of CME and its associated X-ray flare class as input features of the SVM: Accuracy=0.66, PODy=0.76, PODn=0.49, FAR=0.72, Bias=1.06, CSI=0.59, TSS=0.25. The performance of the statistical parameters by applying the SVM is much better than those from the simple classifications based on constant classifiers.

• 한국분무공학회지
• /
• 제25권1호
• /
• pp.8-14
• /
• 2020

Recently, 1-D model-based engine development using virtual engine system is getting more attention than experimental-based engine development due to the advantages in time and cost. Injection rate profile is the one of the main parameters that determine the start and end of combustion. Therefore, it is essential to set up a sophisticated model to accurately predict the injection rate as starting point of virtual engine system. In this research, procedure of 1-D model setup based on AMESim is introduced to predict the dynamic behavior and injection rate of diesel injector. As a first step, detailed 3D cross-sectional drawing of the injector was achieved, which can be done with help of precision measurement system. Then an approximate AMESim model was provided based on the 3D drawing, which is composed of three part such as solenoid part, control chamber part and needle and nozzle orifice part. However, validation results in terms of total injection quantity showed some errors over the acceptable level. Therefore, experimental work including needle movement visualization, solenoid part analysis and flow characteristics of injector part was performed together to provide more accuracy of 1-D model. Finally, 1-D model with the accuracy of less than 10% of error compared with experimental result in terms of injection quantity and injection rate shape under normal temperature and single injection condition was established. Further work considering fuel temperature and multiple injection will be performed.

• Safety and Health at Work
• /
• 제9권2호
• /
• pp.140-143
• /
• 2018

Background: The Occupational Safety and Health Monitoring and Assessment Tool (OSH-MAT) is a practical instrument that is currently used in the German woodworking and metalworking industries to monitor safety conditions at workplaces. The 12-item scoring system has three subscales rating technical, organizational, and personnel-related conditions in a company. Each item has a rating value ranging from 1 to 9, with higher values indicating higher standard of safety conditions. Methods: The reliability of this instrument was evaluated in a cross-sectional survey among 128 companies and its validity among 30,514 companies. The inter-rater reliability of the instrument was examined independently and simultaneously by two well-trained safety engineers. Agreement between the double ratings was quantified by the intraclass correlation coefficient and absolute agreement of the rating values. The content validity of the OSH-MAT was evaluated by quantifying the association between OSH-MAT values and 5-year average injury rates by Poisson regression analysis adjusted for the size of the companies and industrial sectors. The construct validity of OSH-MAT was examined by principle component factor analysis. Results: Our analysis indicated good to very good inter-rater reliability (intraclass correlation coefficient = 0.64-0.74) of OSH-MAT values with an absolute agreement of between 72% and 81%. Factor analysis identified three component subscales that met exactly the structure theory of this instrument. The Poisson regression analysis demonstrated a statistically significant exposure-response relationship between OSH-MAT values and the 5-year average injury rates. Conclusion: These analyses indicate that OSH-MAT is a valid and reliable instrument that can be used effectively to monitor safety conditions at workplaces.