• 대한원격탐사학회지
• /
• 제28권6호
• /
• pp.635-651
• /
• 2012

To produce a level-3 monthly composite image from daily level-2 Sea-viewing Wide Field-of-view Sensor (SeaWiFS) chlorophyll-a concentration data set in the East Sea, we applied four average methods such as the simple average method, the geometric mean method, the maximum likelihood average method, and the weighted averaging method. Prior to performing each averaging method, we classified all pixels into normal pixels and abnormal speckles with anomalously high chlorophyll-a concentrations to eliminate speckles from the following procedure for composite methods. As a result, all composite maps did not contain the erratic effect of speckles. The geometric mean method tended to underestimate chlorophyll-a concentration values all the time as compared with other methods. The weighted averaging method was quite similar to the simple average method, however, it had a tendency to be overestimated at high-value range of chlorophyll-a concentration. Maximum likelihood method was almost similar to the simple average method by demonstrating small variance and high correlation (r=0.9962) of the differences between the two. However, it still had the disadvantage that it was very sensitive in the presence of speckles within a bin. The geometric mean was most significantly deviated from the remaining methods regardless of the magnitude of chlorophyll-a concentration values. Its bias error tended to be large when the standard deviation within a bin increased with less uniformity. It was more biased when data uniformity became small. All the methods exhibited large errors as chlorophyll-a concentration values dominantly scatter in terms of time and space. This study emphasizes the importance of the speckle removal process and proper selection of average methods to reduce composite errors for diverse scientific applications of satellite-derived chlorophyll-a concentration data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권12호
• /
• pp.5928-5947
• /
• 2019

Multicast Mesh based Mobile Ad-hoc NETworks (MANETs) provide efficient data transmission in energy restraint areas without a fixed infrastructure. In this paper, the authors present an improved version of protocol SLIMMER developed by them earlier, and name it SLIMMER-SN. Most mesh-based protocols suffer from redundancy; however, the proposed protocol controls redundancy through the concept of forwarding nodes. The proposed protocol uses remaining energy of a node to decide its energy efficiency. For measuring stability, a new metric called Stability of Node (SN) has been introduced which depends on transmission range, node density and node velocity. For data transfer, a weighted cost function selects the most energy efficient nodes / most stable nodes or a weighted combination of both. This makes the node selection criteria more dynamic. The protocol works in two steps: (1) calculating SN and (2) using SN value in the weighted cost function for selection of nodes. The study compared the proposed protocol, with other mesh-based protocols PUMA and SLIMMER, based on packet delivery ratio (PDR), throughput, end-to-end delay and average energy consumption under different simulation conditions. Results clearly demonstrate that SLIMMER-SN outperformed both PUMA and SLIMMER.

• Nuclear Engineering and Technology
• /
• 제46권2호
• /
• pp.273-280
• /
• 2014

The spectrum weighted responses of various detectors were calculated to provide guidance on the proper selection and use of survey instruments on the basis of their energy response characteristics on the neutron fields. To yield the spectrum weighted response, the detector response functions of 17 neutron-measuring devices were numerically folded with each of the produced calibration neutron spectra through the in-house developed software 'K-SWR'. The detectors' response functions were taken from the IAEA Technical Reports Series No. 403 (TRS-403). The reference neutron fields of 21 kinds with 2 spectra groups with different proportions of thermal and fast neutrons have been produced using neutrons from the $^{241}Am$-Be sources held in a graphite pile, a bare $^{241}Am$-Be source, and a DT neutron generator. Fluence-average energy ($E_{ave}$) varied from 3.8 MeV to 16.9 MeV, and the ambient-dose-equivalent rate [$H^*(10)/h$] varied from 0.99 to 16.5 mSv/h.

• 대한임베디드공학회논문지
• /
• 제15권4호
• /
• pp.197-204
• /
• 2020

The ImageNet dataset is a large scale dataset and contains various natural scene images. In this paper, we propose a convolutional neural network (CNN)-based weighted ensemble technique for the ImageNet classification task. First, in order to fuse several models, our technique uses weights for each model, unlike the existing average-based ensemble technique. Then we propose an algorithm that automatically finds the coefficients used in later ensemble process. Our algorithm sequentially selects the model with the best performance of the validation set, and then obtains a weight that improves performance when combined with existing selected models. We applied the proposed algorithm to a total of 13 heterogeneous models, and as a result, 5 models were selected. These selected models were combined with weights, and we achieved 3.297% Top-5 error rate on the ImageNet test dataset.

• 한국콘텐츠학회논문지
• /
• 제12권2호
• /
• pp.105-115
• /
• 2012

본 논문에서는 휴대폰 카메라를 통해 간판영상의 한글문자를 인식한 후 오인식 된 결과를 교정하는 방법으로 인식 후보를 음소단위 분할하고 연산 가중치를 적용한 weighted Disassemble Levenshtein Distance(wDLD)를 제안한다. 제안된 방법은 인식된 문자열을 음소 단위로 분할한 후 입력 형태의 거리값을 산출하여, 가장 유사한 상호명을 데이터베이스에서 검출 한다. 제안된 방법의 효율성을 검증하기 위해, 전국의 상호명 중 중복되는 상호명을 제거한 130만개의 상호명을 이용하여 데이터베이스 사전을 구축하였다. 또한 대표적인 문자열 비교 알고리즘인 Levenshtein Distance와 음소를 분할하여 적용한 Disassemble Levenshtein Distance 방법, 그리고 본 논문에서 제안한 인식 후보의 음소 단위 분할 방법과 연산 가중치를 적용한 weighted Disassemble Levenshtein Distance의 교정율을 비교 분석 하였다. 그 결과 제안된 weighted Disassemble Levenshtein Distance(wDLD)은 Levenshtein Distance와 Disassemble Levenshtein Distance방법에 비해 각각 평균 29.85%와 6%의 인식률의 향상을 보였다.

• 센서학회지
• /
• 제25권6호
• /
• pp.418-422
• /
• 2016

A filtering algorithm based on the decentralized moving average Kalman filter with uncertainties is proposed in this paper. The proposed filtering algorithm presented combines the Kalman filter with the moving average strategy. A decentralized fusion algorithm with the weighted sum structure is applied to the local moving average Kalman filters (LMAKFs) of different window lengths. The proposed algorithm has a parallel structure and allows parallel processing of observations. Hence, it is more reliable than the centralized algorithm when some sensors become faulty. Moreover, the choice of the moving average strategy makes the proposed algorithm robust against linear discrete-time dynamic model uncertainties. The derivation of the error cross-covariances between the LMAKFs is the key idea of studied. The application of the proposed decentralized fusion filter to dynamic systems within a multisensor environment demonstrates its high accuracy and computational efficiency.

• 한국환경과학회지
• /
• 제17권4호
• /
• pp.397-403
• /
• 2008

The purpose of this study is to investigate weighted compaction density according to a leading density in truck, a compaction density of solid waste and composition ratios of solid waste fur calculation of a capacity of the landfill sites. The experiments for calculations of in-place density at landfill site have been conducted in S landfill site at B City. The size of vessel for measuring the compaction density was $1m^3(1m{\times}1m{\times}1m)$. The experiment tests have been carried out methods (1 time for bulldozer and 4 times for compactor) that do contain all of specification at the landfill site. Average of the loading density at the landfill site was $0.264\;ton/m^3$ ($0.113{\sim}0.487\;ton/m^3$). When the loading density for each compositions was compared, the composition of the highest average loading density ($0.474\;ton/m^3$) was miscellaneous wastes. The composition of the lowest average loading density ($0.120 ton/m^3$) was general solid waste. The reported results indicated that the compaction density at the landfill site was $0.538\;ton/m^3$, which was calculated with weighted incoming ratios of compositions. The ranges of the density for each composition were from $0.021\;ton/m^3$ to $0.221\;ton/m^3$. When the compaction density for each composition was compared, the composition with the highest average compaction density ($0.221\;ton/m^3$) was miscellaneous wastes. The composition with the lowest average compaction density ($0.021\;ton/m^3$) was general solid wastes.

• 한국ITS학회 논문지
• /
• 제2권1호
• /
• pp.63-73
• /
• 2003

국내에서 사용되는 실시간 신호제어시스템은 각 차로별로 계산된 지점속도를 산술평균 모형을 적용하여 접근로의 속도로 추정하고, 이를 대기길이 추정 모형 과정에 중요 변수로 사용한다. 본 연구에서는 현재의 산술모형의 적정성을 검토하고 하나의 대안으로서 교통량 가중평균 모형을 제시하였다. 그리고 교통량 가중평균 모형과평균 개별차랑 지점속도의 이론적 관계를 해석하고, 이를 관계식으로 표현하였다. 제시된 관계식으로부터, 가중평균 모형은 각 차로의 교통량 이용비율과 평균속도 차이에 따라 산술평균과 선형적인 차이를 갖는 것을 파악하였다. 두 가지 모형에 대한 평가를 위하여 실시간 신호제어시스템으로 운영되는 한 개 교차로로부터 자료를 수집하여 각각 적용하였다. 분석 결과, 가중평균 모형이 현재의 산술평균 모형보다 현장 상황을 합리적이고 안정적으로 나타내는 대기 행렬 추정값을 제시하였다. 따라서 가중평균 모형을 실제 운영에 적용하는 것이 실시간 신호제어시스템의 기능적 안정성에 도움을 줄 수 있다고 판단된다.

• 보험의학회지
• /
• 제2권1호
• /
• pp.195-217
• /
• 1985

A statistical analysis on the ratio of weight-for-height in insurance medicine was carried out in a total of 12,690 insured persons who were medically examined at the Honam medical department, Dong Bang life insurance company, ltd. from June, 1979 to February, 1984. The results were as follows: 1. The mean value of the ratio of weight-for-height(W/H ratio) was $102.3{\pm}11.3%$ in all males, and $105.2{\pm}103.6%$ in all females. The difference of the values between males and females showed statistical significance(P<0.001). The mean value of W/H ratio in all the insureds was $103.8{\pm}12.9%$. In each age group, the mean value of female was higher than that of male without exception(P<0.001). The mean value of both sexes in each age group showed gradual increase with age except for the age group over 50(P<0.001). 2. The incidences of the average weighted, overweighted and underweighted lives were respectively 91.6% in all males, 84.2% in all females(P<0.001); 7.9% in all males, 15.4% in all females(P<0.001); 0.5% in all males, 0.4% in all females(P>0.05). The incidences of average weighted, overweighted and underweighted lives in all the insureds were respectively 87.6%, 12.0% and 0.5%. The incidence of the average weighted lives in male in each age group was always higher than that in female with the exception of the age group below 19(P<0.001), whereas the incidence of the overweighted lives in male in each age group was always lower than that in female(P<0.05 in the age group below 19, P<0.001 in the other age groups). The incidence of average weighted lives of both sexes in each age group showed gradual decrease with age from the second to fifth decade(P<0.001), whereas the incidence of the overweighted lives of both sexes in each age group showed gradual increase with age from the second to fifth decade(P<0.001). 3. The relative frequency distribution polygons of W/H ratio of both sexes drawn in a pair on one coordinate plane revealed lying in juxtaposition each other horizontally and showed the shifting of female's polygon to male's one toward the direction of greater value of W/H ratio at a short distance which increased gradually with age. 4. Correlations between both sexes and among age groups relating to W/H ratios of the insureds seem to be a physiological phenomenon of body build and should be considered on the rating of the build. Author prepared a rating table of build based on the correlations above menthioned. Some gradients by sex and age in the rating table of build are organized; in female group, the ratings of 0, 30-50, 50-100 and 100-D are to calculate by the build with W/H ratio of 80-125%, $126{\sim}145%,\;146{\sim}165%$ and over 166% respectively in the age group of 29 or under, by the build with W/H ratio of $80{\sim}130%,\;131{\sim}150%,\;151{\sim}170%$ and over 171% respectively in the age group of $30{\sim}59$, and by the build with W/H ratio of $80{\sim}135%,\;136{\sim}155%,\;156{\sim}175%$ and over 176% respectively in the age group of 60 or over. In male group, the ratings of 0, $30{\sim}50$, 50-100, and 100-D are to calculate by the build with W/H ratio of $80{\sim}120%,\;121{\sim}140%,\;141{\sim}160%$ and over 161% respectively in all ages. Of under weighted lives, in both sexes, the ratings of 30-50 and 50-D are to calculate by the build with W/H ratio of $76{\sim}79%$ and under 75% respectively. The results of the ratings according to a rating table by author were as follows: The incidence of the ratings of 0, 30-50, 50-100 and 100-D showed no difference between male and female. The incidence of the ratings of 0(indicating the average weighted lives) showed gradual decrease with age and the incidence of the ratings of 30-50 and 50-100(indicating the overweighted lives) showed gradual increase with age. These results are thought to be considerably reasonable because the incidence of the builds in each age group is supposed to have no difference by sex and the incidence of the overweighted lives would increase with age.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(1)
• /
• pp.287-290
• /
• 2000

This paper analyze link-sharing mechanisms in packet networks based on the hierarchical class based queueing. The CBQ outlines a set of flexible, efficiently implemented gateway mechanisms that can meet a range of service and link-sharing requirements. We have analyzed the Class level(B, C, D) using the EWMA (Exponential Weighted Moving Average) weight value and EWMA average limit value.