• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제10권1호
• /
• pp.31-36
• /
• 2010

Many clustering algorithms have been used in diverse fields. When we need to group given data set into clusters, many clustering algorithms based on similarity or distance measures are considered. Most clustering works have been based on hierarchical and non-hierarchical clustering algorithms. Generally, for the clustering works, researchers have used clustering algorithms case by case from these algorithms. Also they have to determine proper clustering methods subjectively by their prior knowledge. In this paper, to solve the subjective problem of clustering we make empirical comparisons of popular clustering algorithms which are hierarchical and non hierarchical techniques using Silhouette measure. We use silhouette information to evaluate the clustering results such as the number of clusters and cluster variance. We verify our comparison study by experimental results using data sets from UCI machine learning repository. Therefore we are able to use efficient and objective clustering algorithms.

• ETRI Journal
• /
• 제38권3호
• /
• pp.540-550
• /
• 2016

Spectral clustering is a powerful tool for exploratory data analysis. Many existing spectral clustering algorithms typically measure the similarity by using a Gaussian kernel function or an undirected k-nearest neighbor (kNN) graph, which cannot reveal the real clusters when the data are not well separated. In this paper, to improve the spectral clustering, we consider a robust similarity measure based on the shared nearest neighbors in a directed kNN graph. We propose two novel algorithms for spectral clustering: one based on the number of shared nearest neighbors, and one based on their closeness. The proposed algorithms are able to explore the underlying similarity relationships between data points, and are robust to datasets that are not well separated. Moreover, the proposed algorithms have only one parameter, k. We evaluated the proposed algorithms using synthetic and real-world datasets. The experimental results demonstrate that the proposed algorithms not only achieve a good level of performance, they also outperform the traditional spectral clustering algorithms.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제13권4호
• /
• pp.254-268
• /
• 2013

For real-world clustering tasks, the input data is typically not easily separable due to the highly complex data structure or when clusters vary in size, density and shape. Kernel-based clustering has proven to be an effective approach to partition such data. In this paper, we provide an overview of several fuzzy kernel clustering algorithms. We focus on methods that optimize an fuzzy C-mean-type objective function. We highlight the advantages and disadvantages of each method. In addition to the completely unsupervised algorithms, we also provide an overview of some semi-supervised fuzzy kernel clustering algorithms. These algorithms use partial supervision information to guide the optimization process and avoid local minima. We also provide an overview of the different approaches that have been used to extend kernel clustering to handle very large data sets.

• International Journal of Computer Science & Network Security
• /
• 제23권7호
• /
• pp.109-118
• /
• 2023

Energy awareness is an essential design flaw in wireless sensor network. Clustering is the most highly regarded energy-efficient technique that offers various benefits such as energy efficiency and network lifetime. Clusters create hierarchical WSNs that introduce the efficient use of limited sensor node resources and thus enhance the life of the network. The goal of this paper is to provide an analysis of the various energy efficient clustering algorithms. Analysis is based on the energy efficiency and network lifetime. This review paper provides an analysis of different energy-efficient clustering algorithms for WSNs.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2004년도 추계학술대회
• /
• pp.221-229
• /
• 2004

Recently, there has been enormous growth in the amount of commercial and scientific data, such as protein sequences, retail transactions, and web-logs. Such datasets consist of sequence data that have an inherent sequential nature. However, only a few of the existing clustering algorithms consider sequentiality. This study presents a method for clustering such sequence datasets. The similarity between sequences must be decided before clustering the sequences. This study proposes a new similarity measure to compute the similarity between two sequences using a sequence element. Two clustering algorithms using the proposed similarity measure are proposed: a hierarchical clustering algorithm and a scalable clustering algorithm that uses sampling and a k-nearest neighbor method. Using a splice dataset and synthetic datasets, we show that the quality of clusters generated by our proposed clustering algorithms is better than that of clusters produced by traditional clustering algorithms.

• International Journal of Computer Science & Network Security
• /
• 제22권5호
• /
• pp.17-24
• /
• 2022

Hierarchical clustering methods have been proposed for more than sixty years and yet are used in various disciplines for relation observation and clustering purposes. In 1965, divisive hierarchical methods were proposed in biological sciences and have been used in various disciplines such as, and anthropology, ecology. Furthermore, recently hierarchical methods are being deployed in economy and energy studies. Unlike most clustering algorithms that require the number of clusters to be specified by the user, hierarchical clustering is well suited for situations where the number of clusters is unknown. This paper presents an overview of the hierarchical clustering algorithm. The dissimilarity measurements that can be utilized in hierarchical clustering algorithms are discussed. Further, the paper highlights the various and recent disciplines where the hierarchical clustering algorithms are employed.

• Molecules and Cells
• /
• 제25권2호
• /
• pp.279-288
• /
• 2008

The analysis of microarray data is essential for large amounts of gene expression data. In this review we focus on clustering techniques. The biological rationale for this approach is the fact that many co-expressed genes are co-regulated, and identifying co-expressed genes could aid in functional annotation of novel genes, de novo identification of transcription factor binding sites and elucidation of complex biological pathways. Co-expressed genes are usually identified in microarray experiments by clustering techniques. There are many such methods, and the results obtained even for the same datasets may vary considerably depending on the algorithms and metrics for dissimilarity measures used, as well as on user-selectable parameters such as desired number of clusters and initial values. Therefore, biologists who want to interpret microarray data should be aware of the weakness and strengths of the clustering methods used. In this review, we survey the basic principles of clustering of DNA microarray data from crisp clustering algorithms such as hierarchical clustering, K-means and self-organizing maps, to complex clustering algorithms like fuzzy clustering.

• 정보처리학회논문지B
• /
• 제9B권1호
• /
• pp.7-16
• /
• 2002

웹 검색 엔진의 검색 결과를 클러스터링하는 후처리 클러스터링 알고리즘은 그 특성상 일반적인 클러스터링 알고리즘과는 다른 요구조건을 갖는다. 본 논문에서는 이러한 후처리 클러스터링 알고리즘의 요구조건들을 최대한 만족하는 새로운 클러스터링 알고리즘을 제안하고자 한다. 제안된 Concept ART는 문서 클러스터링에 있어 여러 가지 장점을 갖는 개념 벡터와 실시간 클러스터링 알고리즘으로 알려진 Fuzzy ART를 결합한 형태로써, 후처리 클러스터링뿐 아니라 범용의 클러스터링 알고리즘으로도 응용이 가능하다.

• 한국지능시스템학회논문지
• /
• 제12권3호
• /
• pp.219-227
• /
• 2002

웹 검색 엔진의 검색 결과를 클러스터링하는 후처리 클러스터링 알고리즘은 그 특성상 일반적인 클러스터링 알고리즘과는 다른 요구조건을 갖는다. 본 논문에서는 이러한 후처리 클러스터링 알고리즘의 요구조건들을 최대한 만족하는 새로운 클러스터링 알고리즘을 제안하고자 한다. 제안된 Fuzzy Concept ART는 무서 클러스터링에 있어 여러 가지 장점을 갖는 개념 벡터와 실시간 클러스터링 알고리즘으로 알려진 Fuzzy ART를 퍼지이론에 기반하여 결합한 형태로써, 후처리 클러스터링뿐 아니라 범용의 클러스터링 알고리즘으로도 응용이 가능하다.

• 응용통계연구
• /
• 제26권2호
• /
• pp.361-373
• /
• 2013

군집분석을 위한 알고리즘은 매우 많다. 이러한 군집분석 방법들이 개체들을 어떻게 여러 개의 군집으로 나누는 지를 서로 비교하기 위해서는 나누어지는 군집들이 얼마나 동일한가를 알 수 있는 동의 측도가 필요하다. 우리가 고려하여야 할 군집분석 방법들이 많아질수록 덩달아 동의 측도들 값도 많아지게 된다. 그래서 복수 개의 군집분석 방법들과 대응되는 동의 측도값들을 한 눈에 확인할 수 있는 도구가 필요하다. 본 논문을 통하여 군집분석 방법들과 대응되는 동의 측도값들을 한 눈에 확인할 수 있는 그래픽도구들을 제안하고자 한다.