• 한국컴퓨터정보학회논문지
• /
• 제23권11호
• /
• pp.31-41
• /
• 2018

Text classification has been studied for a long time in the Natural Language Processing field. In this paper, we propose an article- and paragraph-level genre classification system using Word2Vec-based LSTM, GRU, and CNN models for large-scale English corpora. Both article- and paragraph-level classification performed best in accuracy with LSTM, which was followed by GRU and CNN in accuracy performance. Thus, it is to be confirmed that in evaluating the classification performance of LSTM, GRU, and CNN, the word sequential information for articles is better than the word feature extraction for paragraphs when the pre-trained Word2Vec-based word embeddings are used in both deep learning-based article- and paragraph-level classification tasks.

• 품질경영학회지
• /
• 제20권1호
• /
• pp.59-67
• /
• 1992

This paper deals with preventive replacements models for the item whose failures are dependent on their wear level. When measuring the item wear level is very costly, it may be economical to use substitutive characteristics that are correlated with the item wear level and relatively inexpensive to measure. In this paper, replacement policies based on such substitutive characteristics are proposed. The optimal level of substitutive characteristic to replace the item, which minimizes total cost, is obtained. Some numerical examples are also given.

• 전력전자학회:학술대회논문집
• /
• 전력전자학회 2000년도 전력전자학술대회 논문집
• /
• pp.347-350
• /
• 2000

This paper deals with the novel DPWM(discontinuous PWM) for 3-level inverter. Although DPWM methods generate higher harmonics than SVPWM they are of special interest because of their lower switching losses. And in the high modulation region the harmonic characteristics of DPWM is superior to the that of CPWM. However when DPWM applies to the 3-level inverter there is the problem that the output state is varied suddenly in the low modulation region($\textrm{m}_{I}$=0~0.5) The novel DPWM that this problem improves will be introduced.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2008년도 추계학술대회A
• /
• pp.643-645
• /
• 2008

To obtain weight reduction and high performance, level set based topology optimization in magnetic fields is promising for the design of magnetic devices where the precise boundary shape and topological chanages are required. Level set function is introduced to represent ferromagnetic material boundaries and material properties of the magnetic reluctivity are determined. The optimization problem is formulated for maximizing the actuating force in a prescribed direction under limited usage of ferromagnetic material.

• JSTS:Journal of Semiconductor Technology and Science
• /
• 제14권3호
• /
• pp.284-299
• /
• 2014

Coarse-grained reconfigurable architecture (CGRA)-based multi-core architecture aims at achieving high performance by kernel level parallelism (KLP). However, the existing CGRA-based multi-core architectures suffer from much energy and performance bottleneck when trying to exploit the KLP because of poor resource utilization caused by insufficient flexibility. In this work, we propose a new ring-based sharing fabric (RSF) to boost their flexibility level for the efficient resource utilization focusing on the kernel-stream type of the KLP. In addition, based on the RSF, we introduce a novel inter-CGRA reconfiguration technique for the efficient pipelining of kernel-stream on CGRA-based multi-core architectures. Experimental results show that the proposed approaches improve performance by up to 50.62 times and reduce energy by up to 50.16% when compared with the conventional CGRA-based multi-core architectures.

• 한국농공학회논문집
• /
• 제66권2호
• /
• pp.13-24
• /
• 2024

This study aims to contribute to efficient paddy field water management by developing an IoT-based automatic paddy inlet that can consider water level changes according to variations in the supplied water quantity through irrigation channels. This IoT-based automatic paddy inlet not only ensures water level changes based on the supply of irrigation water but also secures irrigation efficiency. The effectiveness and efficiency of the developed IoT-based automatic paddy inlet were presented to contribute to efficient paddy field water management. As a result, the IoT-based automatic paddy inlet demonstrated the capability to maintain the optimal water level in the paddy field. Particularly, it exhibited up to 18.4% higher water resource usage efficiency compared to conventional paddy inlet, emphasizing the IoT-based automatic paddy inlet's advantage in terms of water resource usage.

• Journal of Information Processing Systems
• /
• 제16권6호
• /
• pp.1293-1308
• /
• 2020

In fine-tuning-based transfer learning, the size of the dataset may affect learning accuracy. When a dataset scale is small, fine-tuning-based transfer-learning methods use high computing costs, similar to a large-scale dataset. We propose a mid-level feature extractor that retrains only the mid-level convolutional layers, resulting in increased efficiency and reduced computing costs. This mid-level feature extractor is likely to provide an effective alternative in training a small-scale medical image dataset. The performance of the mid-level feature extractor is compared with the performance of low- and high-level feature extractors, as well as the fine-tuning method. First, the mid-level feature extractor takes a shorter time to converge than other methods do. Second, it shows good accuracy in validation loss evaluation. Third, it obtains an area under the ROC curve (AUC) of 0.87 in an untrained test dataset that is very different from the training dataset. Fourth, it extracts more clear feature maps about shape and part of the chest in the X-ray than fine-tuning method.

• 한국학교보건학회지
• /
• 제9권2호
• /
• pp.171-184
• /
• 1996

The purpose of this study was to investigate how much students stress have, how they cope with this stress, differences between problem behaviors and personality changes based on stress, and stress differing levels when coping with the situation either actively or passively The hypotheses were stated as follows. 1. There will be differences between problem behaviors and personality problems based on sex. 2 There will be differences between problem behaviors and personality problems based on grade. 3. There will be differences between problem behaviors and personality problems based on stress level. 4. There will be differences between problem behaviors and personality problems based on a student's ability to cope with stress. 5. Upper level students under lower stress and lower level students under higher stress will have more problems behaviors and more personality problems. 300 male/female high school students throughout the Seoul area were randomly selected. Of the 300 subjects that were sampled, 294 (Male=145, Female=149) actually participated in this study. The Stress Scale developed by Lazarus & Folkman (1984) Problem Behavior Scale, Personality Problem Scale, and Checklist were used and the conclusions are stated as follows. First, male students have more anti-social behavior and higher anti-social tendencies than female students and female students have more self-depreciation than male students. Second, upper level students have more anti-social behavior, self-ego, fabrication, and higher personality problems than lower level students. Third, students having too much stress have more anti-social behavior, self-ego, and fabrication and personality problems than students having less stress. Fourth, students coping with stress actively showed less self-ego and fabrication and less thought disturbance, anti-social tendencies, and self-depreciation than students coping with stress passively. Finally, upper level students under lower stress and lower level students under higher stress have more fabrication behaviors and more thought disturbance and self-depreciation.

• 한국측량학회지
• /
• 제36권6호
• /
• pp.545-554
• /
• 2018

아날로그 지도에서 디지털 지도로의 지도 패러다임의 변화에 따라 공간정보의 세밀도 개념의 재정의가 필요하다. 이에, 본 연구에서는 디지털 지도 환경에서 활용될 수 있는 4차원 실내 세밀도 모델을 정의하였다. 이를 위하여 기존의 세밀도 개념의 한계점을 도출하고, 이를 기반으로 실내공간정보의 위치 정확도 기반의 위치 세밀도(PLOD: Position accuracy Level Of Detail), 형상 표현기반의 기하 세밀도(GLOD: Geometric Level Of Detail), 일반화 기반의 완성도 세밀도(CLOD: Complete Level Of Detail), 주제 정확도 기반의 의미 세밀도(SLOD: Semantic Level Of Detail)의 4가지의 다른 세밀도를 정의하였다. 또한, 본 연구에서 정의한 4가지의 서로다른 세밀도간의 유기적 관계에 대해 설명하고, 이를 통해 실내 공간정보의 세밀도를 4차원으로 표현하는 방법과 적용 방법 및 예시를 보였다. 향후, 본 연구에서 제시한 4차원의 실내공간 세밀도의 효용성과 타당성을 검증하기 위하여 다양한 실내 서비스를 위한 세밀도 적용 사례 연구와 지형지물 별 완성도 세밀도와 의미 세밀도를 적용하기 위한 연구가 수행되어야 한다.

• Journal of electromagnetic engineering and science
• /
• 제18권3호
• /
• pp.206-211
• /
• 2018

In this paper, we proposed a two-level feature vector fusion technique to improve the performance of target classification. The proposed method combines feature vectors of the early-time region and late-time region in the first-level fusion. In the second-level fusion, we combine the monostatic and bistatic features obtained in the first level. The radar cross section (RCS) of the 3D full-scale model is obtained using the electromagnetic analysis tool FEKO, and then, the feature vector of the target is extracted from it. The feature vector based on the waveform structure is used as the feature vector of the early-time region, while the resonance frequency extracted using the evolutionary programming-based CLEAN algorithm is used as the feature vector of the late-time region. The study results show that the two-level fusion method is better than the one-level fusion method.