• 한국컴퓨터정보학회논문지
• /
• 제26권4호
• /
• pp.21-28
• /
• 2021

본 연구는 기존 편평발 측정을 위해 사용되던 다양한 방법의 한계를 보완할 수 있는 새로운 측정 방법으로 전이학습을 적용한 딥러닝 기반 편평발 분류 방법론을 제안한다. 편평발 88장, 정상발 88장으로 이루어진 총 176장의 이미지 데이터를 활용하여, 적은 데이터로도 우수한 예측 모델을 생성할 수 있는 데이터 증폭 기술과 사전학습 모델인 VGG16 구조를 활용하는 전이학습 기술을 적용하여 제안 모델의 학습을 진행하였다. 제안 모델의 우수성을 확인하기 위하여 기본 CNN 기반 모델과 제안 방법론의 예측 정확도를 비교하는 실험을 수행하였다. 기본 CNN 모델의 경우 훈련 정확도는 77.27%, 검증 정확도는 61.36%, 그리고 시험 정확도는 59.09%로 나타났으며, 제안 모델의 경우 훈련 정확도는 94.32%, 검증 정확도는 86.36%, 그리고 시험 정확도는 84.09%로 나타나 기본 CNN 모델에 비해 제안 모델의 정확도가 큰 폭으로 향상된 것을 확인하였다.

• 한국동물생명공학회지
• /
• 제38권1호
• /
• pp.32-37
• /
• 2023

This study has evaluated the genomic estimated breeding value (GEBV) of the commercial Hanwoo population using the genomic best linear unbiased prediction (GBLUP) method and genomic information. Furthermore, it analyzed the accuracy and realized accuracy of the GEBV. 1,740 heads of the Hanwoo population which were analyzed using a single nucleotide polymorphism (SNP) Chip has selected as the test population. For carcass weight (CWT), eye muscle area (EMA), back fat thickness (BFT), and marbling score (MS), the mean GEBVs estimated using the GBLUP method were 3.819, 0.740, -0.248, and 0.041, respectively and the accuracy of each trait was 0.743, 0.728, 0.737, and 0.765, respectively. The accuracy of the breeding value was affected by heritability. The accuracy was estimated to be low in EMA with low heritability and high in MS with high heritability. Realized accuracy values of 0.522, 0.404, 0.444, and 0.539 for CWT, EMA, BFT, and MS, respectively, showing the same pattern as the accuracy value. The results of this study suggest that the breeding value of each individual can be estimated with higher accuracy by estimating the GEBV using the genomic information of 18,499 reference populations. If this method is used and applied to individual selection in a commercial Hanwoo population, more precise and economical individual selection is possible. In addition, continuous verification of the GBLUP model and establishment of a reference population suitable for commercial Hanwoo populations in Korea will enable a more accurate evaluation of individuals.

• 한국태양에너지학회 논문집
• /
• 제36권6호
• /
• pp.47-59
• /
• 2016

As the usage of sun tracking system in solar energy utilization facility increases, requirement of more accurate computation of sun position has also been increased. Accordingly, various algorithms to compute the sun position have been proposed in the literature and some of them insist that their algorithms guarantee less than 0.01 degree computational error. However, mostly, the true meaning of accuracy argued in their publication is not clearly explained. In addition to that, they do not clearly state under what condition the accuracy they proposed can be guaranteed. Such ambiguity may induce misunderstanding on the accuracy of the computed sun position and ultimately may make misguided notion on the actual sun tracking system's sun tracking accuracy. This work presents some comments related to the implementation of sun position computational algorithm for the sun tracking system. We first introduce the algorithms proposed in the literature. And then, from sun tracking system user's point of view, we explain the true meaning of accuracy of computed sun position. We also discuss how to select the proper algorithm for the actual implementation. We finally discuss how the input factors used in computation of sun position, like time, position etc, affect the computed sun position accuracy.

• The Journal of Asian Finance, Economics and Business
• /
• 제6권4호
• /
• pp.91-98
• /
• 2019

Dividend policy is an important business decision and is considered a channel to communicate a firm's performance to shareholders. Given the empirical findings that earnings quality significantly affects financial analysts' forecasting activities, it is predicted that higher earnings quality would positively influence forecast accuracy. Specifically, it is expected that financial analysts would forecast dividends more accurately for firms with higher earning quality. Unlike the research on financial analysts' earnings forecasts was heavily conducted, there is little study about financial analysts' dividend forecasts. This paper examines the effect of earnings quality on financial analysts' dividend forecast accuracy. We use a sample of South Korean firms for the period of 2011-2015 for multivariate regression. Earnings quality is measured by accruals quality and performance-adjusted discretionary accruals followed by prior studies. We first compare the accuracy between dividend forecasts and earnings forecasts using t-test and Wilcoxon singed-rank test. It is confirmed that financial analysts' dividend forecasts are more accurate than earnings forecasts in Korea. We find that financial analysts' dividend forecasts are more accurate for firms with higher earnings quality. We also find that the result is still valid after controlling for the accuracy of financial analysts' earnings forecasts. This confirms that earnings quality positively affects financial analysts' dividend forecasts.

• Nuclear Engineering and Technology
• /
• 제52권3호
• /
• pp.538-543
• /
• 2020

Boric acid is used as a coolant for pressurized-water reactors, and the degree of burnup is controlled by the concentration of boric acid. Therefore, accurate measurement of the concentration of boric acid is an important factor in reactor safety. An improved system was proposed for the accurate determination of boron concentration. A new boron-concentration measurement technique, called multisensitive detection, was developed to improve the measurement accuracy of boron meters. In previous studies, laboratory-scale experiments were performed based on different sensitivity detectors, confirming a 65% better accuracy than conventional single-detector boron meters. Based on these experimental results, an experimental system simulating the coolant-circulation environment in the reactor was constructed; accuracy analysis of the boron meter with a multisensitivity detector was performed at the actual coolant pressure and temperature. In this study, the boron concentration conversion equation was derived from the calibration test, and the accuracy of the boron concentration conversion equation was examined through a repeatability test. Through the experiment, it was confirmed that the accuracy was up to 87.5% higher than the conventional single-detector boron meter.

• 전자공학회논문지S
• /
• 제35S권6호
• /
• pp.119-126
• /
• 1998

본 논문에서는 인접블럭의 화소단위 움직임벡터와 반화소단위 움직임벡터 사이의 통계적인 특성과 반화소단위 움직임벡터의 공간적 상관성을 이용하여 새로운 반화소단위 움직임벡터 추정 기법들을 제안하였다. 즉, 화소단위 움직임벡터가 같은 인접블럭과는 반화소단위 움직임벡터도 같을 확률이 높다는 사실과 반화소위치들 간에도 서로 인접된 반화소위치간에는 높은 상관성을 지니고 있음을 실험을 통해 확인할 수 있었다. 이러한 성질을 이용한 제안한 방법들은 모의실험 결과 기존의 방법에 비해 아주 미약한 PSNR감소를 보이는 대신 비트 율 및 계산량을 효과적으로 줄일 수 있음을 보여주었다.

• 한국측량학회지
• /
• 제34권2호
• /
• pp.133-142
• /
• 2016

KOMPSAT-3A (KOrea Multi-Purpose SATellite-3A) was launched in March 25 2015 with specification of 0.5 meters resolution panchromatic and four 2.2 meters resolution multi spectral sensors in 12km swath width at nadir. To better understand KOMPSAT-3A positional accuracy, this paper reports a test result on the accuracy of recently released KOMPSAT-3A beta test images. A number of ground points were acquired from 1:1,000 digital topographic maps over the target area for the accuracy validation. First, the original RPCs (Rational Polynomial Coefficients) were validated without any GCPs (Ground Control Points). Then we continued the test by modeling the errors in the image space using shift-only, shift and drift, and the affine model. Ground restitution accuracy was also analyzed even though the across track image pairs do not have optimal convergence angle. The experimental results showed that the shift and drift-based RPCs correction was optimal showing comparable accuracy of less than 1.5 pixels with less GCPs compared to the affine model.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2010년 춘계학술발표회 논문집
• /
• pp.253-255
• /
• 2010

항공사진측량과 위성영상측량에서 거리측정정확도에 대한 연구의 필요성이 급증하고 있다. 그러나 기존 연구들에서는 표준편차와 제곱평균편차간은 물론이고 측정정확도와 측정정밀도간의 정의에 대한 경향성 있는 혼동된 이해가 들어있다. 따라서 본 연구는 항공사진측량과 위성영상측량에서 거리정확도에 관한 표준편차와 제곱평균편차간은 물론이고 측정 정확도와 측정 정밀도간의 관계에 대한 제한적인 정의를 제안한다. 실험결과는 제안한 정확한 정의가 거리측정 정밀도가 아닌 항공사진측량과 위성영상측량에서 거리정확도에서의 개선을 가져옴을 보여준다.

• 한국기계가공학회지
• /
• 제12권2호
• /
• pp.121-126
• /
• 2013

Precision machine tools for high dignity cutting are needed for efforts to improve machining accuracy. However, there are many factors to improve machining accuracy. This study investigated how machining accuracy changes when variation and unbalance amount in rotational speed domain is decreased. Machining accuracy of initial machine tools depends on manufacturing and assembly of parts such as bearing. And then, vibration and noise vary with volume of unbalance amount when it is rotation, so it effects unbalance amount. Also vibration and noise increased by unbalance shorten spindle's life and it especially makes worse boring accuracy. Therefore, this study studied the change of roundness and cylindricity of workpiece when it decreases variation and unbalance in rotational speed domain.

• 대한의용생체공학회:의공학회지
• /
• 제40권2호
• /
• pp.48-54
• /
• 2019

This paper proposed a method of validation data augmentation for improving the grading accuracy of diabetic macular edema (DME) using deep learning. The data augmentation technique is basically applied in order to secure diversity of data by transforming one image to several images through random translation, rotation, scaling and reflection in preparation of input data of the deep neural network (DNN). In this paper, we apply this technique in the validation process of the trained DNN, and improve the grading accuracy by combining the classification results of the augmented images. To verify the effectiveness, 1,200 retinal images of Messidor dataset was divided into training and validation data at the ratio 7:3. By applying random augmentation to 359 validation data, $1.61{\pm}0.55%$ accuracy improvement was achieved in the case of six times augmentation (N=6). This simple method has shown that the accuracy can be improved in the N range from 2 to 6 with the correlation coefficient of 0.5667. Therefore, it is expected to help improve the diagnostic accuracy of DME with the grading information provided by the proposed DNN.