• Journal of the Korea Society of Computer and Information
• /
• v.14 no.3
• /
• pp.183-192
• /
• 2009

ADESS(Adverse drug event surveillance system) is the system which distinguishes adverse drug events using adverse drug signals. This system shows superior effectiveness in adverse drug surveillance than current methods such as volunteer reporting or char review. In this study, we built clinical data mart(CDM) for the development of ADESS. This CDM could obtain data reliability by applying data quality management and the most suitable clustering number(n=4) was gained through the reproducibility assessment in unsupervised learning techniques of knowledge discovery. As the result of analysis, by applying the clustering number(N=4) K-means, Kohonen, and two-step clustering models were produced and we confirmed that the K-means algorithm makes the most closest clustering to the result of adverse drug events.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.4
• /
• pp.179-187
• /
• 2009

Most real world systems contain a series of dynamic and complex phenomena. One of common methods to understand these systems is to build a model and analyze the behavior of them. A two-step methodology comprised of clustering and then model creation is proposed for the analysis on time series data. An interface is designed for CRM(Customer Relationship Management) that provides user with 1:1 customized information using system modeling. It was confirmed from experiments that better clustering would be derived from model based approach than similarity based one. Clustering is followed by model creation over the clustered groups, by which future direction of time series data movement could be predicted. The effectiveness of the method was validated by checking how similarly predicted values from the models move together with real data such as stock prices.

• Convergence Security Journal
• /
• v.10 no.2
• /
• pp.11-17
• /
• 2010

Recent technological advances allow us to envision a future where large numbers of low-power, inexpensive sensor devices are densely embedded in the physical environment, operating together in a wireless network. This paper considers localization for mobile sensors; localization must be invoked periodically to enable the sensors to track their location. Localizing more frequently allows the sensors to more accurately track their location in the presence of mobility. In this paper, we test and analyze the accuracy of a moving node localization by Received Signal Strength (RSS).

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.491-493
• /
• 1998

영상인식은 최근 시각정보의 중요성과 영상을 취득장비의 발달, 처리기술의 향상으로 여러 분야에서 그 중요성과 활용도가 급격히 증가하고 있다. 본 논문에서는 도심지 항공 영상에서 자동표적인식에 관한 문제에서 탐색 물체 주변에 건물들이 밀집되어 있고, 배경이 존재하는 경우에서 fused 분할 방법을 이용하여 기존의 에지 기준 방법인 허프 변환, 에지연결 등에서 발생하는 군집화 문제점을 해결하다. 취득환경의 차이에 다른 농도치 차이를 BCT 방법으로 정규화하여 유사도 기준치로 편차오차를 계산하여 인식하였다. 실험에서는 다양한 탐색물체를 대상으로 회전, 이동, 신축 등의 복합적인 변형에 대하여 불변적으로 인식한 결과를 보였으며, 영상 정합, 컴퓨터 비전, 영상 분석, 영상 이해등의 분야에 적용 가능성을 제시하였다.

• Korean Journal of Plant Resources
• /
• v.35 no.5
• /
• pp.641-651
• /
• 2022

The biodiversity research of barley, a functional food, is proceeding to conserve germplasms and develop new cultivar of barley to improve its functional effects. In this study, with 25,104 barley germplasms in the National Agrobiodiversity Center, South Korea, the biodiversity index of species was much lower (1.17) than the origins (24.73) because of the presence of a biased species, Hordeum vulgare subsp. vulgare, but the species and origin of germplasms were significantly different with regard to genetic traits. In the clustering analysis based on genetic traits, we found that 97% barley germplasms could mostly be distributed between 1~7 clusters out of a total of 15 clusters; 'normal and uzu type', 'lodging', and 'loose smut' were commonly represented in the 1~7 clusters and some clusters showed specific differences in five genetic traits including 'growth habit'. In correlation of each genetic trait, the infection of 'barley yellow mosaic virus' was highly correlated to 'number of grains per spike'. '1000 grain weight' was weakly correlated with seven genetic traits including 'number of grains per spike'. Our analysis for barley's biodiversity can provide a useful guide to the species' phenotypes that need to be collected to conserve biodiversity and to breed new barley varieties.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.4
• /
• pp.1942-1950
• /
• 2013

This paper suggests a new method of detecting attacks of network traffic by visualizing original traffic data and applying multi-class SVM (support vector machine). The proposed method first generates 2D images from IP and ports of transmitters and receivers, and extracts linear patterns and high intensity values from the images, representing traffic attacks. It then obtains variance of ports of transmitters and receivers and extracts the number of clusters and entropy features using ISODATA algorithm. Finally, it determines through multi-class SVM if the traffic data contain DDoS, DoS, Internet worm, or port scans. Experimental results show that the suggested multi-class SVM-based algorithm can more effectively detect network traffic attacks.

• Journal of Korea Multimedia Society
• /
• v.11 no.11
• /
• pp.1501-1514
• /
• 2008

The face shape extracted by the depth values has different appearance as the most important facial information. The face images decomposed into frequency subband are signified personal features in detail. In this paper, we develop a method for recognizing the range face images by multiple frequency domains for each depth image using the modified fuzzy c-mean algorithm. For the proposed approach, the first step tries to find the nose tip that has a protrusion shape on the face from the extracted face area. And the second step takes into consideration of the orientated frontal posture to normalize. Multiple contour line areas which have a different shape for each person are extracted by the depth threshold values from the reference point, nose tip. And then, the frequency component extracted from the wavelet subband can be adopted as feature information for the authentication problems. The third step of approach concerns the application of eigenface to reduce the dimension. And the linear discriminant analysis (LDA) method to improve the classification ability between the similar features is adapted. In the last step, the individual classifiers using the modified fuzzy c-mean method based on the K-NN to initialize the membership degree is explained for extracted coefficient at each resolution level. In the experimental results, using the depth threshold value 60 (DT60) showed the highest recognition rate among the extracted regions, and the proposed classification method achieved 98.3% recognition rate, incase of fuzzy cluster.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.9
• /
• pp.75-84
• /
• 2012

HMM-based Text-to-Speech systems generally utilize context dependent tri-phone units from a large corpus speech DB to enhance the synthetic speech. To downsize a large corpus speech DB, acoustically similar tri-phone units are clustered based on the decision tree using context dependent information. Context dependent information includes phoneme sequence as well as prosodic information because the naturalness of synthetic speech highly depends on the prosody such as pause, intonation pattern, and segmental duration. However, if the prosodic information was complicated, many context dependent phonemes would have no examples in the training data, and clustering would provide a smoothed feature which will generate unnatural synthetic speech. In this paper, instead of complicate prosodic information we propose a simple three prosodic boundary types and decision tree questions that use rising tone, falling tone, and monotonic tone to improve naturalness. Experimental results show that our proposed method can improve naturalness of a HMM-based Korean TTS and get high MOS in the perception test.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11b
• /
• pp.769-771
• /
• 2005

데이터 특성이 연속적이고 애매할 때 퍼지규칙으로 분류 규칙을 표현하는 것은 매우 유용하고 효과적이다. 그러나 일반적으로 정확하지 않은 데이터 특성에 대해서 소속함수를 결정한다는 것은 어려운 일이다. 본 논문에서는 진화알고리즘을 이용하여 효과적인 퍼지 분류 규칙을 자동으로 생성하는 방법을 제안한다. 제안한 방법에서 규칙의 정확성과 이해성을 고려하여 최적화된 소속함수를 생성하기 위해 진화알고리즘을 사용한다. 먼저 지도 군집화로 진화를 위한 초기 소속함수를 생성한다. 진화알고리즘은 전역적 최적 해를 찾는데 효과적이다. 그러나 시간에 대한 효율성이 낮다. 특히 모델 최적화 문제에서는 개체 평가 단계에서 많은 시간이 소요된다. 따라서 본 논문에서는 전체 데이터를 여러 개의 부분 데이터들로 나누고 개체들은 전체 데이터 대신 매번 부분 데이터를 임의적으로 선택하여 개체를 평가함으로써 수행 시간을 단축시킬 수 있는 진화 방법을 제안한다. 제안한 퍼지 분류 규칙 생성 방법의 타당성을 검증하기 위한 실험 데이터로 UCI에서 제공하는 데이터들을 사용하였으며, 실험 결과는 기존 방법에 비해 평균적으로 더 효과적임을 확인하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.562-563
• /
• 2019

본 연구는 한국 프로 야구 선수 개인의 수치화된 데이터를 바탕으로 타석의 결과를 예측하고자 하는데 목적을 두고 있다. 연구의 방법은 2015시즌부터 2018시즌에 활약한 한국 프로 야구 소속의 투수와 타자의 유형을 군집화 하여 지도학습 모델을 만든다. 지도학습 모델과 현재까지 진행된 2019시즌의 결과를 비교·대조한다. 본 연구결과는 한국 프로 야구 10개 구단의 감독의 선수 선발 결정에 기여할 것으로 판단된다.