• ETRI Journal
• /
• 제13권2호
• /
• pp.34-41
• /
• 1991

In this paper, a new distributed parallel algorithm for pattern classification based upon Self-Organizing Neural Network(SONN)[10-12] is developed. This system works without any information about the number of clusters or cluster centers. The SONN model showed good performance for finding classification information, cluster centers, the number of salient clusters and membership information. It took a considerable amount of time in the sequential version if the input data set size is very large. Therefore, design of parallel algorithm is desirous. A new distributed parallel algorithm is developed and experimental results are presented.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2010년도 춘계학술대회
• /
• pp.535-539
• /
• 2010

본 논문은 전파 수집 및 감시 시스템 등에 사용되는 신호의 도래 방향 측정에 관한 연구로써, V/UHF 대역의 다섯 개의 다이폴 안테나를 등 간격으로 원형 배열하여 얻어지는 안테나 간의 위상차 데이터를 CVDF(Correlation Vector Direction Finding) 알고리즘을 적용하여 방향을 추정하였다. CVDF 알고리즘을 적용하여 방향탐지 정확도를 높이기 위해서는 기본적으로 안테나로부터 얻어지는 위상 패턴이 왜곡되지 않고 이상적은 패턴을 가져야 한다. 그러나, 실 환경에서 원형 배열안테나의 위상패턴은 특정 주파수 대역, 특정 방위에서 왜곡되어 나타날 수 있다. 그 이유는 안테나 한 개의 소자만 사용할 때와는 다르게 안테나를 원형 배열구조로 배치함으로써 안테나 각 소자간의 간섭 및 센터폴(원형배열의 중앙에 위치하는 안테나 지지대)의 영향으로 인해 위상 패턴의 왜곡이 생길 수 있다. 특히 안테나 감도를 좋게 하기 위해 신호 증폭 특성을 갖는 Active 안테나를 사용하게 됐을 때는 Passive 안테나를 사용했을 때보다 이런 왜곡현상이 더 크게 나타나게 된다. 본 논문에서는 단순 CVDF 알고리즘을 적용했을 때의 방향탐지 측정 능력이 저하되는 현상을 최소화하여 방향탐지 측정능력 개선시키기 위한 방법으로, 위상을 측정하는 안테나의 조합을 실시간으로 변경하는 방법 및 안테나 빔패턴을 활용하는 방법등을 제안한다. 위의 제안된 개선알고리즘 적용 전/후의 시험결과를 통하여 이 제안방안의 타당성을 확인한다.

• 인터넷정보학회논문지
• /
• 제17권6호
• /
• pp.53-59
• /
• 2016

최근 무선 센서 네트워크, 사물 인터넷, 소셜 네트워크 서비스와 같은 다양한 응용 분야에서 대용량 스트림 데이터가 실시간으로 생성되고 있으며, 효율적인 기법을 통해 처리 및 분석하여 유용한 정보를 찾아내고, 이를 의사 결정을 위해 사용할 수 있도록 하는 것은 중요한 이슈 중에 하나이다. 스트림 데이터는 끊임없이 빠른 속도로 생성되므로 최소한의 접근을 통해 처리해야 하며, 신속한 저전력 처리를 필요로 하는 자원이 제한된 환경에서 분석될 수 있도록 적합한 기법이 요구된다. 이러한 문제를 해결하기 위해, 슬라이딩 윈도우 개념이 제안되어 연구되고 있다. 한편, 대용량 데이터로부터 의미 있는 정보를 찾아내기 위한 데이터 마이닝 기법 중에 하나인 패턴 마이닝은 중요 정보를 패턴 형태로 추출한다. 전통적인 빈발 패턴 마이닝은 이진 데이터베이스를 대상으로 하고 모든 아이템을 동일한 중요도로 고려함으로써 데이터 마이닝 분야에서 중요한 역할을 수행해 왔지만, 실제 데이터 특성을 반영하지 못하는 단점을 지닌다. 하이 유틸리티 패턴 마이닝은 비 이진 데이터베이스로부터 상대적인 아이템 중요도를 반영하여 더욱 의미 있는 정보를 찾아내기 위해 제안되었다. 정적 데이터를 대상으로 하는 하이 유틸리티 패턴 마이닝 기법은 그러나 스트림 데이터 처리에 적합하지 못하다. 제한된 환경에서 스트림 데이터의 특성을 반영하고 효율적으로 처리하여 중요한 정보를 찾아내기 위해 슬라이딩 윈도우 기반의 접근법이 제안되었다. 본 논문은 슬라이딩 윈도우 기반 하이 유틸리티 패턴 마이닝 기법들의 성능을 평가하고 분석하여 해당 기법들의 특성 및 발전 방향을 고찰한다.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 2006년도 추계학술대회
• /
• pp.427-430
• /
• 2006

Clustering for gene expression data without filtering out noise genes may be distorted or derived inappropriate inference. Identifying significant genes and deleting noise before major analysis is necessary fur meaningful discovery from genes expression pattern. We proposed a new method of finding significant genes using factor analysis which is done on transposed data matrix. We construct significance score that is sum of factor loadings for declared significant number of factor, and set threshold through replication. Our proposed method works well for simulated time-course data for finding significant genes even though variance level gets larger.

• 한국공간구조학회논문집
• /
• 제18권2호
• /
• pp.25-34
• /
• 2018

This study investigated a bio-tensegrity structural system that combines the characteristics of a general tensegrity structural system with a biological system. The final research objective is to accomplish a changeability for the structural system as like the movement of the natural bio-system. In the study, we present a shape finding procedure for the two stage bio-tensegrity system model inspired by the movement pattern of animal backbone. The proposed system is allowing a dynamic movement by introducing the concept of "saddle" for the variable bio-tensegrity structure. Several shape finding analysis example and results are presented and shows a efficient validation and suitability.

• 대한가정학회지
• /
• 제8권1호
• /
• pp.21-35
• /
• 1970

1. For the purpose of making the basic pattern construction 100 girls attending the Seoul National University College of Home Economics were measured in finding of body size and coefficient of correlation which would be used as one basis of this study. 2. Coefficient of correlation of each body part based on the breast width was shown as follows; Correlation coefficients of bust to waist and hip were high and those of bust to shoulder width, neck height, back width, and breast width were low. None of that was found between bust and back length. It was not recognized so scientific to adjust the basic pattern construction with figures proportioned by those of neck, shoulder width, breast width, and back width. 3. The method of basic pattern construction obtained by this research has been demonstrated in direct wearing since 1967. It is believed that the result will contribute a great benefit in teaching clothing as well as in mass production industry of ready-made garments.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1992년도 하계학술대회 논문집 A
• /
• pp.462-464
• /
• 1992

In this paper, an extraction of depth intonation in pattern using neural network is presented. All the 3D images represent the depth information in grey pixels. This pixels which have analog values translated digital values. Because of the noise and distortion in pattern, we use the normalization in learning and recalling the patterns. Our method has eight direction vectors and slopes for pattern. Also, we use potential to obtain the mean slope and direction vectors of given 3D patches. The higher level of deduction finding the global depth information is also carried out by using neural network.

• 대한한방부인과학회지
• /
• 제29권2호
• /
• pp.113-120
• /
• 2016

Objectives: This study aims to report the effects of Korean medicine treatment by pattern identification and Sweet Bee Venom Pharmacopuncture on Chronic Relapsing Cystitis Methods: The patients was treated with Korean medicine by pattern identification and Sweet Bee Venom Pharmacopuncture at Qugu (CV2), Guanyuan (CV4). We evaluated treatment effects by changes of symptoms and urine analysis (UA) finding. Results: After treatments, the clinical symptoms such as painful urination, dysuria, frequent urination were improved and the state of urinalysis was improved. Conclusions: This clinical study suggests that Korean medicine treatment by pattern identification and Sweet Bee Venom Pharmacopuncture are significantly effective in treatment of a Chronic Relapsing Cystitis

• 한국주거학회:학술대회논문집
• /
• 한국주거학회 2005년도 추계학술대회 논문집
• /
• pp.311-314
• /
• 2005

Our city structure is being complicated from the development of industry and the increase in population. It makes the ground space overcrowded, so underground space is systematized for use. Because this underground space has very complicated configuration, people have difficulty in finding the route in the underground space. Therefore It is demanded to comprehend configuration of the various space structure, predict pattern of the moving and analyze induction factor of the moving. When people are in the conflagration or dangerous situation, they are in the habit of using the route which they are used to go to and from. It is needed to analyze moving pattern of the pedestrian who usually use the underground space for the sake of dealing with conflagration in the underground space. Finally, this study suggested that general characteristic, distribution, moving route of the pedestrian and factors influencing that they choose the route which the research finds out.

• 한국언어정보학회지:언어와정보
• /
• 제9권1호
• /
• pp.69-92
• /
• 2005

Pattern recognition in cognitive agents is based on (i) the uninterpreted input data (e.g. parameter values) provided by the agent's hardware devices and (ii) and interpreted patterns (e.g. templates) provided by the agent's memory. Computationally, the task consists in finding the memory data corresponding best to the input data, for any given input. Once the best fitting memory data have been found, the input is recognized by applying to it the interpretation which happens to be stored with the memorized pattern. This paper presents a fast converging procedure which starts from a few initially recognized items and then analyzes the remainder of the input by systematically checking for items shown by memory to have been related to the initial items in previous encounters. In this way, known patterns are tried first, and only when they have been exhausted, an elementary exploration of the input is commenced. Efficiency is improved further by choosing the candidate to be tested next according to frequency.