• The Korean Journal of Applied Statistics
• /
• v.20 no.1
• /
• pp.167-181
• /
• 2007

Many clustering algorithms and cluster validation techniques for high-dimensional gene expression data have been suggested. The evaluations of these cluster validation techniques have, however, seldom been implemented. In this paper we compared various cluster validity indices for low-dimensional simulation data and real gene expression data, and found that Dunn's index is the most effective and robust, Silhouette index is next and Davies-Bouldin index is the bottom among the internal measures. Jaccard index is much more effective than Goodman-Kruskal index and adjusted Rand index among the external measures.

• 한국데이터정보과학회:학술대회논문집
• /
• 2006.04a
• /
• pp.63-76
• /
• 2006

Several clustering algorithms to analyze gene expression data and cluster validation techniques that assess the quality of their outcomes, have been suggested, but evaluations of these cluster validation techniques have seldom been implemented. In this paper we compared various cluster validity indices for simulation data and real genomic data, and found that Dunn's index is more effective and robust through small simulations and with real gene expression data.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.675-677
• /
• 2005

This study proposes a fuzzy classification using EM algorithm. For cluster validation, this approach iteratively estimates the class-parameters in the fuzzy training for the sample classes and continuously computes the log-likelihood ratio of two consecutive class-numbers. The maximum ratio rule is applied to determine the optimal number of classes.

• Industrial Engineering and Management Systems
• /
• v.8 no.2
• /
• pp.101-108
• /
• 2009

Biclustering is a method to extract subsets of objects and features from a dataset which are characterized in some way. In contrast to traditional clustering algorithms which group objects similar in a whole feature set, biclustering methods find groups of objects which have similar values or patterns in some features. Both in clustering and biclustering, validating how much the result is informative or reliable is a very important task. Whereas validation methods of cluster solutions have been studied actively, there are only few measures to validate bicluster solutions. Furthermore, the existing validation methods of bicluster solutions have some critical problems to be used in general cases. In this paper, we review several well-known validation measures for cluster and bicluster solutions and discuss their limitations. Then, we propose several improved validation indices as modified versions of existing ones.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.9 no.6
• /
• pp.621-626
• /
• 1999

In this paper a new cluster validation index which is heuristic but able to eliminate the monotonically decreasing tendency occurring in which the number of cluster c gets very large and close to the number of data points n is proposed. We review the FCM algorithm and some conventional cluster validity criteria discuss on the limiting behavior of the proposed validity index and provide some numerical examples showing the effectiveness of the proposed cluster validity index.

• The Korean Journal of Applied Statistics
• /
• v.19 no.1
• /
• pp.13-32
• /
• 2006

Biologists are attempting to group genes based on the temporal pattern of gene expression levels. So far, a number of methods have been proposed for clustering microarray data. However, the results of clustering depends on the genes selection, therefore the gene selection with significant expression difference is also very important to cluster for microarray data. Thus, this paper present the results of broad comparative studies to time course microarray data by considering methods of gene selection, clustering and cluster validation.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.171-174
• /
• 2005

기존의 클러스터 평가 지표(cluster validation index)는 클러스터의 개수가 커질수록 클러스터 평가 지표 값이 단조 감소하는 경향을 보인다. 최근에 이러한 단점을 보완하는 새로운 클러스터 평가 지표가 본 논문 저자중의 하나에 의해 제안되었으나, over-clustering의 단점 을 지니고 있다. 본 논문에서는, 클러스터 평가 지표 값이 단조 감소 및 over-clustering을 방지할 수 있는 새로운 클러스터 평가 지표를 제안하고, 여러 가지 예제를 통하여 새롭게 제안된 평가 지표의 타당성을 보인다.

• Korean Journal of Remote Sensing
• /
• v.20 no.5
• /
• pp.353-359
• /
• 2004

This study proposes an approach to unsupervisedly estimate the number of classes and the parameters of defining the classes in order to train the classifier. In the proposed method, the image is segmented using a spatial region growing based on hierarchical clustering, and fuzzy training is then employed to find the sample classes that well represent the ground truth. For cluster validation, this approach iteratively estimates the class-parameters in the fuzzy training for the sample classes and continuously computes the log-likelihood ratio of two consecutive class-numbers. The maximum ratio rule is applied to determine the optimal number of classes. The experimental results show that the new scheme proposed in this study could be used to select the regions with different characteristics existed on the scene of observed image as an alternative of field survey that is so expensive.

• Journal of KIISE:Software and Applications
• /
• v.31 no.12
• /
• pp.1591-1601
• /
• 2004

Clustering, a technique for the analysis of the genes, organizes the patterns into groups by the similarity of the dataset and has been used for identifying the functions of the genes in the cluster or analyzing the functions of unknown gones. Since the genes usually belong to multiple functional families, fuzzy clustering methods are more appropriate than the conventional hard clustering methods which assign a sample to a group. In this paper, a Bayesian validation method is proposed to evaluate the fuzzy partitions effectively. Bayesian validation method is a probability-based approach, selecting a fuzzy partition with the largest posterior probability given the dataset. At first, the proposed Bayesian validation method is compared to the 4 representative conventional fuzzy cluster validity measures in 4 well-known datasets where foray c-means algorithm is used. Then, we have analyzed the results of Saccharomyces cell cycle expression data evaluated by the proposed method.

• Journal of Navigation and Port Research
• /
• v.48 no.2
• /
• pp.78-87
• /
• 2024

To quantitatively assess the correlation between subsystems within a port cluster and the overall coordinated development of the port group, the current paper evaluates the coordinated development of port clusters. First, we construct an evaluation index system for the coupling and coordination of port clusters. Next, we introduce the contribution index of port subsystems, coupling degree, and coupling coordination degree functions to formulate a coupling coordination evaluation model for the port cluster. Finally, we use the Yangtze River Delta port cluster as a case study for validation, specifically using empirical data from 2012 to 2021. The findings reveal distinct phased characteristics in the coupling and coordination of port clusters in the Yangtze River Delta, marked by a notable transition from "maladjustment" to "coordination." Further, sustained high coupling values over a decade indicate a significant level of competition and cooperation among ports within the Yangtze River Delta port cluster. Over time, this competitive and collaborative dynamic has progressively evolved toward a more positive and structured direction. Lastly, it is expected that the evaluation model proposed in this paper can be extrapolated to other port clusters to gauge the extent of coordinated development, thereby facilitating horizontal comparisons and vertical analyses.