• 한국농업기계학회:학술대회논문집
• /
• 한국농업기계학회 2017년도 춘계공동학술대회
• /
• pp.166-166
• /
• 2017

모돈의 건강 상태를 정량 지수화 하기 위한 연구를 수행 중이다. 지제이상, 섭식 불량, 수면 패턴 등의 운동 특성 분석을 위하여 복수의 초음파 센서를 이용하였다. 시계열 계측 신호를 분석하여 정량 지수화를 수행하는 과정에서 주파수 도메인 분석을 시도하였다. 이 과정에서 주파수 도메인의 분해능에 따른 편차 극복을 위한 비선형 모델링을 수행하였다. 또한 인접한 시계열 데이터 구간 간의 상관성 분석이 가능하면 대용량 데이터의 실시간 처리로 인한 지연 시간 극복 및 기대되는 예후에 대한 조기 진단이 가능할 것이다. 본 연구에서는 구글에서 제공하는 Tensorflow와 NVIDIA에서 제공하는 CUDA 엔진을 동시 적용한 심층 학습 시스템을 이용하였다. 전 처리를 위하여 주파수 분해능 (2분, 3분, 5분, 7분, 11분, 13분, 17분, 19분)에 따른 데이터 집합을 1단계로 두고, 상위 10 순위 안에 드는 파워 스펙트럼 밀도의 크기를 2단계로 하여, 총 2~10개의 입력 노드를 순차적으로 선정하였고, 동일한 방식으로 인접한 시계열의 파워 스펙터럼 밀도를 순위를 변화시켜 지정하였다. 대표적인 심층학습 모델인 Softmax regression with a multilayer convolutional network를 이용하여 Recursive feature selection 경우의 수를 $8{\times}9{\times}9$로 총 648 가지 선정하고, Epoch는 10,000회로 지정하였다. Calibration 모델링의 경우 Cost function이 10% 이하인 경우 해당 경우의 학습을 중단하였으며, 모델 간 상호 교차 검증을 수행하기 위하여 $_8C_2{\times}_8C_2{\times}_8C_2$ 경우의 수에 대한 Verification test를 수행하였다. Calibration 과정 상 모든 경우에 대하여 10% 이하의 Cost function 값을 보였으나, 검증 테스트 과정에서 모든 경우에 대하여 $r^2$ < 0.5 인 결정 계수 값이 나타났다. 단적으로 심층학습 모델의 과도한 적합(Over fitting) 방식의 한계를 보인 것이라 판단할 수 있다. 적합한 Feature selection 및 심층 학습 모델에 대한 지속적이고 추가적인 고려를 통해 과도적합을 해소함과 동시에 실효적이고 활용 가능한 Classification을 위한 입, 출력 노드 단의 전후 Indexing, Quantization에 대한 고려가 필요할 것이다. 이를 통해 모돈 생체 정보 정량화를 위한 지능형 현장 진단 기술 연구를 지속할 것이다.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1994년도 추계학술대회
• /
• pp.43-46
• /
• 1994

A new spatial localization technique using high-order surface gradient coil (SGC) is proposed. Although the Spatial Selection with High-Order gradient (SHOT) can provide a 2-D selection with only one selective RF pulse, the high-order gradient produced by cylindrical-shape coils has not been clinically useful for clinical systems due to the large minimum selection size caused by the limited radial gradient intensity. However, by using the proposed high-order SGCs located near the imaging region, the size of volume selection can be reduced to a clinically useful 1-4 cm in diameter by applying stronger radial gradient with much less gradient driving power. A 40 cm-by-40 cm $r^{2}$ SGC has been designed and constructed, and phantom and volunteer studies have been performed. Experimental results using spatially localized MRI show good agreement to the theoretically predicted behavior.

• International Journal of Industrial Entomology and Biomaterials
• /
• 제10권2호
• /
• pp.95-99
• /
• 2005

A breeding plan was carried out on four commercial strains of silkworm (Bombyx mori L.) 101, 102, 103 and 104 to improve some important traits. Genetic gain or response to selection $({\Delta}G)$, heritability of cocoon shell weight (CSW) and specific combining ability effects were estimated to determine the strains that can be improved. Strain 101 had lowest heritabitity, ${\Delta}G$ and viability. Strain 102 was acceptable in selection response but its viability was low. Therefore these two strains were not suitable for more selection. As a result, only lines 103 and 104 were chosen for further improvement. Intra population selection based on independent culling level method practiced from third to sixth generation for both productive and viability traits simultaneously. While CSW and CW had increasingly enhanced during primary generations, they went slightly up after third generation. According to negative genetic correlation, viability decreased during primary generations, but after third generation that paid attention to balanced development of both productive and viability traits, viability increased so that the pupation rate reached to $91\%$ in 103 and $97\%$ in 104 for last generation $(G_8)$.

• ETRI Journal
• /
• 제41권3호
• /
• pp.358-370
• /
• 2019

The main aim of this study is to select the optimal set of genes from microarray cancer datasets that contribute to the prediction of specific cancer types. This study proposes the enhancement of the feature selection filter algorithm based on Joe's normalized mutual information and its use for gene selection. The proposed algorithm is implemented and evaluated on seven benchmark microarray cancer datasets, namely, central nervous system, leukemia (binary), leukemia (3 class), leukemia (4 class), lymphoma, mixed lineage leukemia, and small round blue cell tumor, using five well-known classifiers, including the naive Bayes, radial basis function network, instance-based classifier, decision-based table, and decision tree. An average increase in the prediction accuracy of 5.1% is observed on all seven datasets averaged over all five classifiers. The average reduction in training time is 2.86 seconds. The performance of the proposed method is also compared with those of three other popular mutual information-based feature selection filters, namely, information gain, gain ratio, and symmetric uncertainty. The results are impressive when all five classifiers are used on all the datasets.

• 천문학회보
• /
• 제41권2호
• /
• pp.64.2-64.2
• /
• 2016

Optical variability is one way to probe the nature of the central engine of AGN at smaller linear scales and previous studies have shown that optical variability is more prevalent at longer timescales and at shorter wavelengths. Especially, intra-night variability can be explained through the damped random walk model but small samples and inhomogeneous data have made constraining this model hard. To understand the properties and physical mechanism of optical variability, we are performing the KMTNet Active Nuclei Variability Survey (KANVaS). Test data of KMTNet in the COSMOS field was obtained over 2 separate nights during 2015, in B, V, R, and I bands. Each night was composed of 5 and 9 epochs with ~30 min cadence. To find AGN in the COSMOS field, we applied multi-wavelength selection methods. Different selection methods means we are looking different region in unification model of AGN, and 100~120, 400~500, 50~100 number of AGN are detected in X-ray, mid-infrared, and radio selection of AGN, respectively. We performed image convolution to reflect seeing fluctuation, then differential photometry between the selected AGN and nearby stars to achieve photometric uncertainty ~0.01mag. We employed one of the standard time-series analysis tools to identify variable AGN, chi-square test. Preliminarily results indicate that intra-night variability is found for X-ray selected, Type1 AGN are 23.6%, 26.4%, 21.3% and 20.7% in the B, V, R, and I band, respectively. The majority of the identified variable AGN are classified as Type 1 AGN, with only a handful of Type 2 AGN showing evidence for variability. The work done so far confirms that there are type and wavelength dependence of intra-night optical variability of AGN.

• 수산해양기술연구
• /
• 제59권3호
• /
• pp.189-195
• /
• 2023

This study quantitatively evaluated size selectivity for three netting shapes (T0; regular, T45, T90) and hanging ratio (35%, 70%) of T0 netting used for trawl codend. The size selectivity experiment was performed in a tank using a cube experimental model with a length of 50 cm on one side and 389 experimental individuals, jack mackerel (Trachurus japonicus). In the selectivity analysis, a selectivity curve was created based on the selection ratio using a logistic function, and the 25%, 50%, and 70% selection length and selection range (SR) were obtained. The T0 netting was 19.54 cm when the 50% selective length, which is a selectivity evaluation index, had a hanging ratio of 35%, a selection range of 0.51 cm, and 22.70 cm and 3.08 cm for the hanging ratio of 70%. The T45 netting was 24.34 cm and 2.13 cm, and the T90 netting was 23.51 cm and 2.84 cm. The results of the T45 netting and the T90 netting are similar, and the 50% selection length and selection range were relatively larger than the T0 netting. There was a significant difference in the correlation between the circumference of the inner circle of the mesh by the shape of the netting and the body girth of the experimental individual (Pearson test, r = 0.86, p < 0.05). There was no significant difference in the correlation between the selection ratio by the T0 netting, T45 netting, and T90 netting with a 70% hanging ratio (one-way ANOVA, p > 0.05). The results of this study showed that selectivity such as T45 netting and T90 netting appeared when the hanging ratio, which maximizes the area of T0 netting, was maintained at 70%.

• 디지털융복합연구
• /
• 제8권2호
• /
• pp.203-216
• /
• 2010

본 연구는 디지털화된 한국인 인체정보 데이터세트의 전략적 활용계획수립과 관련된다. 구체적으로 국가경쟁력 제고를 위한 R&D 전략 포트폴리오의 작성을 위하여 필요한 중요한 의사결정 요인들을 파악하고 조직화하는 것이다. 타 국가의 경우를 보더라도 디지털 인체정보 데이터세트와 시각화 에프리케이션의 개발은 국가수준에서 전략적인 R&D 프로젝트로 선정하여 추진하고 있다. 본 연구에서는 연구목적을 달성하기 위하여 해당 분야의 전문가 집단을 구성하였으며, 이들을 통하여 R&D 비전, SWOT분석 및 전략개발, 연구분야 및 세부과제를 파악하였다. 또한 전략계획 수립을 위하여 각 세부과제들을 중요도와 시급도 관점에서 우선순위를 파악하였다. 연구의 방법으로는 브레인스토밍, 델파이방법과 AHP(Analytical Hierarchy Process) 기법을 도입하였다. 본 연구의 결과는 추후 R&D 포트폴리오 작성을 위한 가이드라인이 될 뿐 아니라, 해당 분야에 연구투자를 평가하는 프레임워크로서 중요한 역할을 담당할 것이다.

• 대한산업공학회지
• /
• 제3권2호
• /
• pp.63-72
• /
• 1977

The trend of computerization is significant in Korea even at its beginning stage, especially for governmental administration. However, full-fledged success of computerization in an organization is reported to be rare while the cost of computerization has been high and increasing. This paper is concerned with two features for the successful implementation of a computerized system in an organization selection criteria for the computerization among the possible candidate projects and project selection models. Due to the dearth of literature regarding successful implementation of computerization, other sources of literature with respect to R & D management, method engineering, etc. were reviewed to develop a set of factors influencing successful computerization. Thus, project selection criteria for computerization of governmental administration are developed and organized as follows: cost of computerization project including both system development and operating cost, quanitative and qualitative benefits of computerization project, probability of technical and implementation success of computerization and other organizational and political factors to be considered. These criteria are broken down into detailed sets of subcriteria to be measured. To select a project after thorough consideration of the selection criteria, a project selection model which takes into account all criteria together has to be developed. In the study three project selection models are suggested and developed. They are the checklist model, multi-stage cut-off model, and composite criteria model. A detailed procedure for each of the three models is illustrated. Although the project selection criteria and models are developed here primarily for the computerization of governmental administration, they are easily applicable to other settings of computerization. Finally, some caveats for the use of selection criteria and models are discussed.

• Asian-Australasian Journal of Animal Sciences
• /
• 제17권2호
• /
• pp.159-163
• /
• 2004

Data pertaining to first lactation records of 316 Murrah buffaloes, progeny of 47 sires, maintained at NDRI Farm for a period of 18 years were analysed to construct selection indices and to examine their robustness by changing the relative economic values of different economic traits. A total of 120 selection indices were constructed for three sets of relative economic values ( 40 for each set) considering different combinations of seven first lactation traits viz. age at first calving (AFC), first lactation 305 day or less milk yield (FLMY), first lactation length (FLL), first calving interval (FCI), milk yield per day of first lactation length (MY/FLL), milk yield per day of first calving interval (MY/FCI) and milk yield per day age at second calving (MY/ASC). The three sets of relative economic values were based on economic values of different traits, 1% standard deviation of different traits and regression of different traits on FLMY. The 'optimum' indices for the first two sets had five traits each namely AFC, FLMY, FLL, FCI and MY/ASC giving improvement in aggregate genotype of Rupees 269.11 and Rs. 174.88, respectively. The accuracy of selection from both indices was 70.79 and 69.39%, respectively. The 'best' selection index from the third set of data again had five traits (AFC, FLMY, FLL, FCI and MY/FLL) giving genetic gain of Rs. 124.16 and accuracy of selection of 71.81%. The critcal levels or break-even points for FLMY for varying levels of AFC and FCI estimated from the "optimum index" suggested the need of enhancement of present production level of the herd or reduction of AFC or FCI. It was concluded that economic values of various first lactation traits were the most appropriate to construct selection indices as compared to other criteria of assigning relative economic weights in Murrah buffaloes.

• Bulletin of the Korean Chemical Society
• /
• 제32권7호
• /
• pp.2397-2404
• /
• 2011

Variable selection k nearest neighbor QSAR modeling approach was applied to a data set of 80 3-arylisoquinolines exhibiting cytotoxicity against human lung tumor cell line (A-549). All compounds were characterized with molecular topology descriptors calculated with the MolconnZ program. Seven compounds were randomly selected from the original dataset and used as an external validation set. The remaining subset of 73 compounds was divided into multiple training (56 to 61 compounds) and test (17 to 12 compounds) sets using a chemical diversity sampling method developed in this group. Highly predictive models characterized by the leave-one out cross-validated $R^2$ ($q^2$) values greater than 0.8 for the training sets and $R^2$ values greater than 0.7 for the test sets have been obtained. The robustness of models was confirmed by the Y-randomization test: all models built using training sets with randomly shuffled activities were characterized by low $q^2{\leq}0.26$ and $R^2{\leq}0.22$ for training and test sets, respectively. Twelve best models (with the highest values of both $q^2$ and $R^2$) predicted the activities of the external validation set of seven compounds with $R^2$ ranging from 0.71 to 0.93.