Earticle

With the conventional computational fluid dynamics (CFD) approaches like unstructured high-density mesh method, the problem about the solution dependencies on the grid and on the physical models isn't completely resolved. In this study, thus, a new approach named Building-Cube Method based on a Cartesian mesh is proposed using a high-density mesh in order to solve the problem of the current CFD approaches. In the present method, a flow field is divided into a number of cubes (squares in 20) of various sizes. Each cube is a computational sub-domain with Cartesian mesh of equal spacing and equal number of nodes. The geometrical size of each cube is determined by adapting to geometry and flow features so as to cope with broadband characteristic lengths of the flow. Equal spacing and equal number of Cartesian mesh in each cube make it easy to parallelize the flow solver and to handle huge data output. The method is applied to several airfoils including NACA0012 and two-element airfoils at relatively low Reynolds number, RAE2822 airfoil with transition trip at high Reynolds number and a four-element airfoil at high Reynolds number. These results validate the capability of the present approach.

In this study, the effects of the compaction method for (Mg+2B) powders on the apparent density and superconducting properties of MgB2 bulk superconductor were investigated. The raw powders used in this study were nano-sized boron (B) and spherical magnesium (Mg). A batch of a powder mixture of (Mg+2B) was put in a steel mold and uniaxially pressed at 1 ton or 3 tons into pellets. Another batch of the powder mixture was uniaxially pressed at 1 ton and then pressed isostatically at 1800 kg/cm2 in the water chamber. All pellets were heat-treated at 650℃ for 1 h in flowing argon gas for the formation of MgB2. The apparent density of powder compacts pressed at 3 ton was higher than that at 1 ton. The cold isostatic pressing (CIP) in a water chamber allowed further increase of the apparent density of powder compacts, which influenced the pellet density of the final products (MgB2). The compaction methods (uniaxial pressing and CIP) did not affect the formation of MgB2 and superconducting critical temperature (Tc) of MgB2, but affected the critical current density (Jc) of MgB2 significantly. The sample with the high apparent density showed high Jc at 5 K and 20 K at applied magnetic fields (0-5 T).

현대 신경망은 종종 높은 분류 정확도를 달성하지만, 여전히 신뢰도 보정 오류 문제를 겪고 있어 의사결정이 중요한 응용 분야에서 신뢰성이 제한된다. 본 논문에서는 신경망 신뢰도의 신뢰성을 향상시키기 위해 MADCAT이라 불리는 2단계 사후 보정 방법을 제안한다. 제안된 방법은 먼저 로짓 마진을 샘플 난이도의 지표로 사용하여 샘플별 온도를 예측하고, 차선 레이어 특징 공간에서 K-최근접 이웃 거리를 기반으로 한 밀도 인식 조정을 통해 보정된 신뢰도를 정교화 한다. 또한, 최적화와 보정 품질 간의 정합성을 높이기 위해 훈련 목표로 소프트-빈 기대 보정 오차를 채택한다. CIFAR-10, CIFAR-100 및 ImageNet에 대한 실험결과, 제안된 방법은 컨볼루션 신경망과 비전 트랜스포머 모두에서 뛰어난 보정 성능을 달성하는 동시에, 클래스 불균형, 분포 이동 및 제한된 검증 데이터 환경에서도 미미한 추론 오버헤드로 효과적인 성능을 유지하는 것으로 나타났다.

Modern neural networks often achieve high classification accuracy but still suffer from confidence miscalibration, which limits their reliability in decision-sensitive applications. In this paper, a two stage post-hoc calibration method, called MADCAT, is proposed to improve the reliability of neural network confidence. The proposed method first uses the logit margin as an indicator of sample difficulty to predict a sample-specific temperature, and then refines the calibrated confidence through a density-aware adjustment based on K-Nearest Neighbor distances in the penultimate-layer feature space. In addition, Soft-Binned Expected Calibration Error is adopted as the training objective to better align optimization with calibration quality. Experimental results on CIFAR-10, CIFAR-100, and ImageNet show that the proposed method achieves strong calibration performance on both convolutional neural networks and Vision Transformers, while remaining effective under class imbalance, distribution shift, and limited validation data settings with negligible inference overhead.

Forest inventory is based on the sampling methods which estimates the population statistic in terms of calculating stand volumes, and so the researches on cost-effective, precise, and convenient method of computing stand volumes are still important issues in South Korea. The aims of this study were to analyze the applicability of point sampling methods by comparing the complete enumeration and square plot sampling and to figure out the number of points per ha for increasing the precision with regard to the density and percentage of sample points. Study area was 9.25 ha with two sub-compartment (4.3 ha and 4.85 ha) and mainly covered by Japanese larch (Larix kaempferi (Lamb.) Carriere) and natural deciduous forests, located in the research forest of Kangwon National University. Point sampling was carried out with 2 and 4 basal area factor (BAF) and the number of points were 134, which were sampled in the grid unit of 30 m × 30 m. Square plots were sampled with 21 plots in the grid unit of 20 m × 20 m. The stand basal area was 22.9 m2/ha and the stand density was 507 trees/ha in complete enumeration. These statistics were similar to the results of using point sampling: 21.1 m2/ha for stand basal area and 509 trees/ha for stand density. However, it was quite different from the results of using square plots: 38.9 m2/ha for stand basal area and 700 trees/ha for stand density. To analyze the density and percentage of point sampling, the number of plots were categorized into five groups based on the 134 plots (100%): 10 plots (7.5%), 20 plots (14.9%), 40 plots (29.9%), 60 plots (44.8%), 100 plots (74.6%). As a result of analyzing the basal area and stand density randomly, the difference occurred when 10 to 20 plots were used, but there was no large difference between 40 and 100 plots. Assuming that the area of forest inventory which conducts in South Korea is generally 5 to 30 ha, the number of sampling points in this study were different from that in the previous studies, and thus, it is considered to be used for the practical use.

Cultural practices of Asparagus cochinchinensis in highland area were performed for a potential medicinal crop. These studies were examined to propagation methods, planting densities, nitrogen treatments, and cropping years. The results are summarized as follows. The adequate number of buds per tuberous root was 4 for vegetative propagation because the number of tuberous root harvested was 16.8 and the yield was also the highest, exhibiting 1,060 ㎏/10a. The suitable planting time for vegetative propagation was later than early April. If the earlier tuberous roots were planted, the less they emerged. The highest emergence rate was obtained from the planting density of 30 × 20㎝ as 97.2% while the yield was highest in the 30 × 15㎝ density, exhibiting 1,883 ㎏/10a with emergence rate at 94.9. It seemed that the higher planting density promoted plant height growth and yield in Asparagus cochinchinensis. The highest fresh weight was recorded at 6 ㎏/10a of nitrogen fertilizer into the sandy loam soil compared to the level of 0, 3, 9 ㎏/10a. The yield was increased with cropping years. However, the proper harvesting time was the second year of cultivation because the rate of weight increase was maximized in the 2-year-old tuberous root. The yield in the third year was decreased as compared to that of the second year.

PURPOSES : This paper presents a mix design method for using steel slag as an aggregate for asphalt mixtures. METHODS : Steel slag has a different density and absorption rate than natural aggregates. The asphalt content was calculated according to the steel slag characteristics, and the formula for aggregate-gradation correction was presented. RESULTS : The asphalt mix was designed using the proposed equations. Using the proposed mix design method, it was possible to design the asphalt mixture according to the target-usage amount of the recycled aggregate. CONCLUSIONS : The suggested method can be used for asphalt mix design using aggregates with different densities and absorption rates. It is expected to contribute to quality improvement by ensuring accurate calculation of mixing ratios for steel slag asphalt mixtures.

In ultrasonic bone densitometry, the positioning of measurement site is decisive in precision and reproducibility. In this study, automatic Region of Interest (ROI) detection algorithm is suggested and adopted the method using the local minimum value by ultrasonic image. The preprocess before the local minimum method extracts out the bone area and calculates the geometrical information of bone. The developed ROI detection algorithm was applied to the clinical test for the subject of 305 female patients in the range of 22-88 years old. As the results, the accuracy of the algorithm was shown to be 98.3％. It was also found that bone density parameter was significantly correlated with age(r=0.85, p<0.0001).

우리는 도래각 (DoA; direction of arrival) 추정 방법 중 하나인 서로소 배열 기반의 프로퍼게이터 방법을 개선시키는 알고리즘을 제안한다. 서로소 배열 기반의 프로퍼게이터 방법은 특이값 분해없이 도래각을 추정하는 방법으로 서로소 배열 기반의 MUSIC에 비하여 현저히 낮은 복잡도를 지녔으나, 다소 저하된 도래각 추정 성능을 보인다. 우리는 이러한 성능 저하의 원인 중 하나로 잡음의 파워 스펙트럼 밀도를 포함하고 있는 신호의 자기상관행렬의 대각 성분이 사용되고 있지 않음에 있음을 파악하고, 잡음의 파워 스펙트럼 밀도가 장기간에 걸쳐 추정이 가능하다는 사실에 착안하여 신호의 자기상관행렬의 대각 성분을 사용하는 도래각 추정 방법을 제안한다. 우리는 시뮬레이션을 통해 우리가 제안한 방법이 기존의 서로소 배열 기반의 프로퍼게이터 방법보다 연산량을 4배정도 증가시키지만 탐지확률 95% 기준 하에 신호대 잡음비를 1.5dB, 도래각 분해능을 $0.7^{\circ}$ 만큼 개선시켜 그 성능이 서로소 배열 기반의 MUSIC에 보다 근접함을 관찰한다.

We propose an improved direction of arrival (DoA) estimation algorithm based on co-prime array and propagator method. The propagator method with co-prime array does not require singular value decomposition (SVD) requiring much less computational complexity but exhibiting somewhat worse performance in comparison with MUSIC based on co-prime array. We notice that one cause of the performance degradation was in the avoidance of the usage of the diagonal elements of the signal autocorrelation matrix that contains the noise power spectral density. So we propose an algorithm with the diagonal elements of the signal autocorrelation matrix based on the fact that the noise power spectral density can be estimated using noise observation over a long period of time. We observe, through simulations, that the proposed scheme in this paper improves the performance, with 4 times more computational requirement, by signal-to-noise ratio of 1.5dB and by DoA resolution of $0.7^{\circ}$ at the detection probability of 95% compared with the previously introduced co-prime array propagator scheme, resulting in performance much closer to that of co-prime array-based MUSIC scheme.

Since domestic offset print quality control depends on the experience of persons in charge rather than objective data, the standards from the data are needed for the well-organized quality management of high-quality prints. Based on the ISO standards, now the standard specifications, such as GRACoL 7, SWOP 11, SNAP, G7 in America, and EURO Color, ECI, BVDM in Europe, and Japan Color 2001, Japan Color 2004 in Japan, are employed. There exists KS for domestic standards. However, since it has been directly imported from ISO Standards, there are discrepancies in printing materials and media. Although the quality management of commercial prints has been focused and studied. For this circumstance, standards are needed for the production of prints to meet international standards. This paper aims to find out suitable ways to approach domestic printing standards using densitometric method and colorimetric method. Printing experiments on this study were performed with domestic process color inks(Cyan, Magenta, Yellow, Black) and domestic coated papers, under the same conditions as the actual commercial printing processes. In desitometric method, how to set up the optimal solid print density was studied through three tests such as the ways to match solid print density and print contrast to GRACoL standards respectively, and the way to get the highest contrast. In colorimetric method, how to get solid print density with the minimum color difference between the CIEL*a*b* color specification values in GRACoL standards and the measured color specification values in prints, was studied.

Moisture content of printing paper and circumstance of printing room has great influence on the printing works. We have studied on the printability of domestic art paper according to the insufficient Korea weather. As the result, we found the printability of domestic art paper are much more affect by moisture content than any other printability coefficients, especially print density and hygroexpansivity decrease with moisture content increase.

In urban planning, center identification has been one of the most important subjects in terms of urban growth management and efficiency. Therefore, many studies have investigated more accurate and efficient methodology in order to identify the urban center such as employment density, network concentration density and land price gradient. They have used the administrative boundary as the spatial analysis unit because almost all data are collected by it. However, it has serious limitation in detecting the accurate boundary of urban center because it is restricted within the existing administrative unit boundary. This study suggests the Density-Based Clustering (DENCLUE) method as the alternative which identifies more accurate urban center boundary. We examined pros and cons of DENCLUE method on the Gangnam subcenter identification with 9-Dong employment data. The results showed that DENCLUE can improve the accuracy of center identification and expectation power about future growth of center. In fact, the center identification results by DENCLUE are significantly different from those by the existing methods which use administrative unit.

This research attempts on plate making for gravure. Direct Hard Dot Method, Helio Kliso Graph, and Laser TB Dot; these three method are the most popular in Korea. For examining ink transmission, three plates were made in the above three different way for the same manuscript and were printed under the same condition. After printing speed was also varied into 150line/inch, 175line/inch, and 200line/inch, ink transmission was examined too. After printing the same manuscript with the above three method then the researcher examined level of ink transmission. Printing lines shows that Laser TB Dot Method and Direct Hard Dot Method were favorable level, while there was big differences of state of Helio Kliso Graph plate.

거시적인 3대 교통변수(교통량, 속도, 밀도)와 고속도로 서비스 수준의 효과척도(밀도, 교통량 대 용량비) 모두 해당되는 변수는 밀도이다. 특히 도로의 서비스 수준을 평가함에 있어 가장 정확하고 우수한 효과척도로 알려져 있다. 이러한 중요성에도 불구하고 측정방법의 어려움으로 인해 타 변수에 비해 밀도를 활용한 연구가 상대적으로 부족하였다. 기존 밀도추정방법의 경우 측정시각에 따라 동일한 교통류의 밀도값이 변화하는 등의 한계가 있다. 본 연구에서는 서울외곽순환고속도로 사패산 터널 내부의 CCTV 영상을 정합하여 파노라마 영상을 제작한 후, 제작된 파노라마 영상을 이용하여 실제 밀도를 측정하는 방법을 연구하였다. 중심극한정리를 이용하여 분석한 결과, 1 km 사진 24개(혹은 24초)를 이용하여 밀도를 측정하면 당시 교통상황을 잘 반영할 수 있었다. 즉, 본 연구에서 제시한 밀도 수집 주기를 준수하여 수집한 구간영상으로 밀도를 측정할 경우 측정시각에 관계없이 평균적인 밀도값을 측정할 수 있으며, 현실 교통류를 대변하는 현실적인 밀도를 취득할 수 있다.

Density is applied both three major macroscopic traffic variables (traffic volume, speed, and density) and two measures of effectiveness (MOE) for level of service (LOS) on highway (density and V/C). Especially, it is known for the most accurate MOE on evaluating the LOS of highway. Despite such importance, there is a lack of study on density relatively than other variables for its difficulty of measurement. Existing density estimation methods have some limitations such as density values of same traffic flow vary with collecting time. In this study, we researched actual density measuring method with panoramic image, after each CCTV images in the Sapaesan Tunnel on Seoul Ring Expressway are matched into one panoramic image. Analysis through the Central Limit Theorem shows that density of 24 1 km-images, which means 24 second, applies traffic situation well. That is to say that reasonable density value regardless of collecting time, and practical density which represents actual traffic flow can be taken in case of measuring density by suggested collecting cycle.

본 연구에서는 옵션가격에 내재된 위험중립확률을 추정하는 Shimo(1993)의 방법에 대해 몇 가지 개선사항을 제시한다. 아울러 이 방법을 적용하여 우리나라 KOSPI 200 주가지수옵션 시장에서 위험중립확률을 추출하고 이것의 형태가 역사적으로 어떻게 변화되어 왔는지를 살 펴보고자 한다. 본 연구는 기존의 연구와 다음의 측면에서 차별화된다. 첫째, 거래빈도가 많은 외가격과 등가격옵션의 가격을 최대한 반영하고 거래가 많지 않은 심내가격(deep ITM) 옵션의 가격 을 상대적으로 적게 반영하기 위하여 거래량의 제곱근을 가중치로 사용하여 옵션의 내재변 동성 함수와 이에 따른 위험중립확률을 추정하였다. 거래가 많은 옵션일수록 그 옵션의 가 격에 시장에 대한 투자자들의 기대감이 충분히 더 많이 반영되어 있을 것이기 때문이다. 둘 째, 옵션의 내재변동성을 행사가격의 함수로 도출하기 위해 비선형 최소제곱근사(least squares fitting)를 사용해야 하는데, 이러한 문제의 고질적인 특성인 해의 불안정성을 극복 하기 위하여 본 연구에서는 새로운 알고리즘을 적용해 보았다. 마지막으로 본 연구에서는 KOSPI 200 주가지수 옵션가격으로부터 추출된 위험중립확률분포가 2002년 이후 주식시장 의 부침에 따라 어떻게 변화되었는지를 고찰하였다.

Workers’ anomalous trajectories allow us to detect emergency situations in the workplace, such as accidents of workers, security threats, and fire. In this work, we develop a scheme to detect abnormal trajectories of workers using the edit distance on real sequence (EDR) and density method. Our anomaly detection scheme consists of two phases: offline phase and online phase. In the offline phase, we design a method to determine the algorithm parameters: distance threshold and density threshold using accumulated trajectories. In the online phase, an input trajectory is detected as normal or abnormal. To achieve this objective, neighbor density of the input trajectory is calculated using the distance threshold. Then, the input trajectory is marked as an anomaly if its density is less than the density threshold. We also evaluate performance of the proposed scheme based on the MIT Badge dataset in this work. The experimental results show that over 80 % of anomalous trajectories are detected with a precision of about 70 %, and F1-score achieves 74.68 %.

Norovirus has become the most common cause of human gastroenteritis in developed countries. Detection procedures of foodborne viruses from foods require several steps. The concentration step using polyethylene glycol (PEG) is time-consuming and the detection efficiency of reverse transcription-polymerase chain reaction (RT-PCR) is affected by inhibitors from food components. In this study, a rapid detection method based on buoyant density centrifugation was developed to replace the time-consuming chloroform-polyethylene glycol-Tris·Tween method. Feline calicivirus that belongs to the family Caliciviridae was used as a surrogate model for norovirus. After artificial inoculation of feline calcivirus (FCV) to oyster and lettuce, 830 μL of homogenized sample suspension was layered on the top of 670 μL 20% percoll and centrifuged. Then RNA extraction step was proceeded with the supernatant. By varying several physical conditions, the detection limits were lowered to 2.4×102 PFU per 1 g in oyster and 2.4×100 PFU per 1 g in lettuce. The protocol obtained in this study could be used to develop new detection method for norovirus in foods.

Density-based spatial clustering of applications with noise (DBSCAN) is a base algorithm for density based clustering. It can find out the clusters of different shapes and sizes from a large amount of data, which is containing noise and outliers. However, it fails to handle the local density variation that exists within the cluster. Thus, a good clustering method should allow a significant density variation within the cluster because, if we go for homogeneous clustering, a large number of smaller unimportant clusters may be generated. In this paper an enhancement of DBSCAN algorithm is proposed, which detects the clusters of different shapes, sizes that differ in local density. We introduce new algorithm Dynamic Method DBSCAN (DMDBSCAN). It selects several values of the radius of a number of objects (Eps) for different densities according to a k-dist plot. For each value of Eps, DBSCAN algorithm is adopted in order to make sure that all the clusters with respect to the corresponding density are clustered. For the next process, the points that have been clustered are ignored, which avoids marking both denser areas and sparser ones as one cluster. Experimental results are obtained from artificial data sets and UCI real data sets. The final results show that our algorithm get a good results with respect to the original DBSCAN and DVBSCAN algorithms.

In the computer-aided medical diagnosis, the problem of missing attribute values in many medical data sets brings a great challenge to data mining. To solve the problem, this paper proposes a method based on density clustering and grey relational analysis. It provides an effective solution for missing medical data. The method uses the characteristic and degree of data samples dynamic relation and the existing attribute value information to impute the missing value, in order to alleviate the difficulty brought by missing data for aided medical diagnosis. By comparison with the experiment, the proposed method can effectively solve the classification problem which causes by missing medical attribute value, accurately predict the patient’s health and provide the help to doctor’s diagnosis.