• 한국지능시스템학회논문지
• /
• 제13권6호
• /
• pp.661-666
• /
• 2003

본 논문에서는 범주형 데이터의 분류를 위한 새로운 기법을 제시한다. 기존의 대표적인 퍼지 군집화 방법인 k-modes 알고리즘과 fuzzy k-modes 알고리즘은 군집의 중심을 단일 값으로 표현하고, 군집에 속하는 데이터의 빈도 수에 기반한 중신 갱신 기법을 사용하였다. 이와 같은 기존의 방법들은 분류의 경계가 모호한 데이트를 군집화할 경우, 알고리즘의 각 단계에서 발생하는 분류의 에러를 보정하지 못해 최종적으로 지역해에 빠지는 단점이 있다. 이를 극복하기 위해 본 논문에서는 군집 중심을 퍼지 집합을 이용하여 정의한다. 퍼지 군집 중심은 주어진 데이터와 군집간의 거리 관계를 퍼지 값을 이용해 표현하며, 각 군집의 중심은 데이터의 소속 정도 값을 이용해 갱신된다. 이와 같은 퍼지 중심 표현기법을 도입하여 범주형 데이터의 분류 시에 보다 세밀한 결정을 내림으로써, 인접한 군집들의 경계에서 발생하는 불확실성을 최소화한다. 기존의 대표적인 방법들과의 비교실험을 수행함으로써 제안한 방법의 성능을 검증하였다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국지능시스템학회 2007년도 추계학술대회 학술발표 논문집
• /
• pp.423-426
• /
• 2007

클러스터링은 주어진 데이터 포인트들을 주어진 개수의 그룹으로 나누는 비지도 학습의 한 방법이다. 클러스터링의 방법 중 하나로 널리 알려진 퍼지 클러스터링은 하나의 포인트가 모든 클러스터에 서로 다른 정도로 소속될 수 있도록 함으로써 각 포인트가 하나의 클러스터에만 속할 수 있도록 하는 K-means와 같은 방법에 비해 자연스러운 클러스터 형태의 유추가 가능하고, 잡음에 강한 장점이 있다. 이 논문에서는 기존의 퍼지 클러스터링 방법 중 소속도(membership)와 전형성(typicality)을 동시에 계산해 낼 수 있는 Possibilistic Fuzzy C-Means (PFCM) 방법에 Gath-Geva (GG)의 방법 을 적용하여 PFCM을 확장한다. 제안한 방법은 PFCM의 장점을 그대로 가지면서도, GG의 거리 척도에 의해 클러스터들 사이의 경계를 강조함으로써 분류 목적에 적합한 소속도를 계산할 수 있으며, 전형성은 가우스 형태의 분포에서 생성된 포인트들의 분포 함수를 정확하게 모사함으로써 확률 밀도 추정의 방법으로도 사용될 수 있다. 또한 GG 방법은 Gustafson-Kessel 방법과 달리 클러스터에 포함된 포인트의 개수가 확연히 차이 나는 경우에도 정확한 결과를 얻을 수 있다는 사실을 실험 결과를 통해 확인할 수 있었다.

• 한국정밀공학회지
• /
• 제20권11호
• /
• pp.47-54
• /
• 2003

Research during the past several years has established the effectiveness of acoustic emission (AE)-based sensing methodologies for machine condition analysis and process. AE has been proposed and evaluated for a variety of sensing tasks as well as for use as a technique for quantitative studies of manufacturing process. STD11 has been known as difficult-to-cut materials. The micro-grooving machine was developed for this study and the experiments were performed using CBN blade for machining STD11. Evaluating the machining conditions, frequency spectrum analysis of acoustic emission (AE) signals according to each conditions were applied. Fuzzy clustering method for associating the preprocessor outputs with the appropriate decisions was followed by frequency spectrum analysis. FFT is used to decompose AE signal into different frequency bands in time domain, the root mean square (RMS) values extracted from the decomposed signal of each frequency band were used as features.

• 전자공학회논문지B
• /
• 제31B권9호
• /
• pp.150-158
• /
• 1994

In this paepr, Fuzzy Constraint Line Clustering (FCLC) method for optical flow estimation is proposed. FCLC represents the spatical and temporal gradients as fuzzy sets. Based on these sets, several constraint lines with different membership values are generated for the poxed whose velocity is to be estimated. We describe the process for obtaining the membership values of the spatial and temporal gradients and that of the corresponding constraint line. We also show the process for deciding the tightest cluster of point formalated by intersection between constraint lines. For the synthetic and real images, the results of FCLC are compared with of CLC.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2011년도 제43차 동계학술발표논문집 19권1호
• /
• pp.81-84
• /
• 2011

본 논문은 전압 이벤트 현상 중 순간전압강하(Sag) 현상에 초점을 맞추었다. Sag 현상의 심각한 정도를 표현하는 심각도(Voltage Sag Severity) 지수는 동일 지속시간에 대한 임계치와의 비로 표현하였다. 제안하는 확장된 심각도(Expanded Severity) 지수는 sag현상의 분포에 따른 일시반복성의 정보를 표현하였다. 기존의 임계치를 표현하는 ITIC curve를 기반으로 된 심각도와 sag 현상이 발생하는 지속시간-전압 그래프의 분포를 fuzzy clustering을 통하여 medoid를 측정하고, medoid의 심각도와 실제 임계치에 근접한 sag 지점의 심각도를 계산하여 비교하였다. 확장된 심각도 지수는 심각도가 높은 현상들과의 연계성을 나타내는 지수로 심각한 정도의 수치 정보 이외에 일시적인 현상인지 지속 반복적인 현상인지를 0과 1사이의 수치로 표현하였고, 실험을 통하여 입증하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권4호
• /
• pp.1692-1711
• /
• 2016

Energy is a scarce resource in wireless sensor networks (WSNs). A variety of clustering protocols for WSNs, such as the zone-based stable election protocol-enhanced (ZSEP-E), have been developed for energy optimization. The ZSEP-E is a heterogeneous zone-based clustering protocol that focuses on unbalanced energy consumption with parallel formation of clusters in zones and election of cluster heads (CHs). Most ZSEP-E research has assumed probabilistic election of CHs in the zones by considering the maximum residual energy of nodes. However, studies of the diverse CH election parameters are lacking. We investigated the performance of the ZSEP-E in such scenarios using a fuzzy logic approach based on three descriptors, i.e., energy, density, and the distance from the node to the base station. We proposed an efficient ZSEP-E scheme to adapt and elect CHs in zones using fuzzy variables and evaluated its performance for different energy levels in the zones.

• Journal of Electrical Engineering and Technology
• /
• 제7권6호
• /
• pp.884-890
• /
• 2012

Hysteretic behavior model of soft magnetic material usually used in electrical machines and electronic devices is necessary for numerical solution of Maxwell equation. In this study, a new dynamic hysteresis model is presented, based on the nonlinear dynamic system identification from measured data capabilities of fuzzy clustering algorithm. The developed model is based on a Gustafson-Kessel (GK) fuzzy approach used on a normalized gathered data from measured dynamic cycles on a C core transformer made of 0.33mm laminations of cold rolled SiFe. The number of fuzzy rules is optimized by some cluster validity measures like 'partition coefficient' and 'classification entropy'. The clustering results from the GK approach show that it is not only very accurate but also provides its effectiveness and potential for dynamic magnetic hysteresis modeling.

• Nuclear Engineering and Technology
• /
• 제42권4호
• /
• pp.414-425
• /
• 2010

In this paper, a procedure is developed for identifying a number of representative solutions manageable for decision-making in a multiobjective optimization problem concerning the test intervals of the components of a safety system of a nuclear power plant. Pareto Front solutions are identified by a genetic algorithm and then clustered by subtractive clustering into "families". On the basis of the decision maker's preferences, each family is then synthetically represented by a "head of the family" solution. This is done by introducing a scoring system that ranks the solutions with respect to the different objectives: a fuzzy preference assignment is employed to this purpose. Level Diagrams are then used to represent, analyze and interpret the Pareto Fronts reduced to the head-of-the-family solutions.

• 전기학회논문지
• /
• 제59권5호
• /
• pp.981-989
• /
• 2010

In this study, we introduce a new architecture of fuzzy inference system. In the fuzzy inference system, we use Fuzzy C-Means clustering algorithm to form the premise part of the rules. The membership functions standing in the premise part of fuzzy rules do not assume any explicit functional forms, but for any input the resulting activation levels of such radial basis functions directly depend upon the distance between data points by means of the Fuzzy C-Means clustering. As the consequent part of fuzzy rules of the fuzzy inference system (being the local model representing input output relation in the corresponding sub-space), four types of polynomial are considered, namely constant, linear, quadratic and modified quadratic. This offers a significant level of design flexibility as each rule could come with a different type of the local model in its consequence. Either the Least Square Estimator (LSE) or the weighted Least Square Estimator (WLSE)-based learning is exploited to estimate the coefficients of the consequent polynomial of fuzzy rules. In fuzzy modeling, complexity and interpretability (or simplicity) as well as accuracy of the obtained model are essential design criteria. The performance of the fuzzy inference system is directly affected by some parameters such as e.g., the fuzzification coefficient used in the FCM, the number of rules(clusters) and the order of polynomial in the consequent part of the rules. Accordingly we can obtain preferred model structure through an adjustment of such parameters of the fuzzy inference system. Moreover the comparative experimental study between WLSE and LSE is analyzed according to the change of the number of clusters(rules) as well as polynomial type. The superiority of the proposed model is illustrated and also demonstrated with the use of Automobile Miles per Gallon(MPG), Boston housing called Machine Learning dataset, and Mackey-glass time series dataset.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 학술대회 논문집 정보 및 제어부문
• /
• pp.340-342
• /
• 2004

In this paper, we introduce a new category of fuzzy inference systems baled on information granulation to carry out the model identification of complex and nonlinear systems. Informal speaking, information granules are viewed as linked collections of objects(data, in particular) drawn together by the criteria of proximity, similarity, or functionality. Granulation of information with the aid of Hard C-Means(HCM) clustering algorithm help determine the initial parameters of fuzzy model such as the initial apexes of the membership functions and the initial values of polyminial functions being used in the premise and consequence part of the fuzzy rules. And the initial parameters are tuned effectively with the aid of the genetic algorithms(GAs) and the least square method. The proposed model is contrasted with the performance of the conventional fuzzy models in the literature.