• Speech Sciences
• /
• v.10 no.4
• /
• pp.159-166
• /
• 2003

To reduce the high dimensionality required for training of feature vectors in speaker identification, we propose an efficient GMM based on local PCA with Fuzzy clustering. The proposed method firstly partitions the data space into several disjoint clusters by fuzzy clustering, and then performs PCA using the fuzzy covariance matrix in each cluster. Finally, the GMM for speaker is obtained from the transformed feature vectors with reduced dimension in each cluster. Compared to the conventional GMM with diagonal covariance matrix, the proposed method needs less storage and shows faster result, under the same performance.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.671-672
• /
• 2008

K-means is a popular one in clustering algorithms, and it minimizes the mutual euclidean distance among the sample points. But K-means has some demerits, such as depending on initial condition, unsupervised learning and local optimum. However mahalanobis distancecan deal this case well. In this paper, the author proposed a new clustering algorithm, named exponential probability clustering, which applied Mahalanobis distance into K-means clustering. This new clustering does possess not only the probability interpretation, but also clustering merits. Finally, the simulation results also demonstrate its good performance compared to K-means algorithm.

• Journal of the Korean Society for information Management
• /
• v.21 no.3
• /
• pp.251-267
• /
• 2004

The purpose of this study is to generate the local level knowledge structure of a single document, similar to end-of-the-book indexes and table of contents of printed material through the use of term clustering and cluster representative term selection. Furthermore, it aims to analyze the functionalities of the knowledge structure. and to confirm the applicability of these methods in user-friend1y information services. The results of the term clustering experiment showed that the performance of the Ward's method was superior to that of the fuzzy K -means clustering method. In the cluster representative term selection experiment, using the highest passage frequency term as the representative yielded the best performance. Finally, the result of user task-based functionality tests illustrate that the automatically generated knowledge structure in this study functions similarly to the local level knowledge structure presented In printed material.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.13 no.4
• /
• 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.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1995.06a
• /
• pp.83-86
• /
• 1995

Many image display devices allow only a limited number of colors to be simultaneously displayed. In displaying of natural color image using color palette, it is necessary to construct an optimal color palette and map each pixel of the original image to a color palette with fast. In this paper, we proposed the clustering algorithm using local region block centered one color cluster in the prequantized 3-D histogram. Cluster pairs which have the least distortion error are merged by considering distortion measure. The clustering process is continued until to obtain the desired number of colors. Same as the clustering process, original color image is mapped to palette color via a local region block centering around prequantized original color value. The proposed algorithm incorporated with a spatial activity weighting value which is smoothing region. The method produces high quality display images and considerably reduces computation time.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.6
• /
• pp.1230-1237
• /
• 2009

In this paper, we develop a design methodology of Granular-based Radial Basis Function Neural Networks(GRBFNN) by context-based clustering. In contrast with the plethora of existing approaches, here we promote a development strategy in which a topology of the network is predominantly based upon a collection of information granules formed on a basis of available experimental data. The output space is granulated making use of the K-Means clustering while the input space is clustered with the aid of a so-called context-based fuzzy clustering. The number of information granules produced for each context is adjusted so that we satisfy a certain reconstructability criterion that helps us minimize an error between the original data and the ones resulting from their reconstruction involving prototypes of the clusters and the corresponding membership values. In contrast to "standard" Radial Basis Function neural networks, the output neuron of the network exhibits a certain functional nature as its connections are realized as local linear whose location is determined by the values of the context and the prototypes in the input space. The other parameters of these local functions are subject to further parametric optimization. Numeric examples involve some low dimensional synthetic data and selected data coming from the Machine Learning repository.

• Journal of Internet Computing and Services
• /
• v.13 no.5
• /
• pp.1-8
• /
• 2012

In this paper, we propose a new clustering combination based on numerical data driven for rule generation mechanism. In large and complicated space, a clustering method can obtain limited performance results. To overcome the single clustering method problem, hybrid combined methods can solve problem to divided simple cluster estimation. Fundamental structure of the proposed method is combined by mountain clustering and modified Chen clustering to extract detail cluster information in complicated data distribution of non-parametric space. It has automatic rule generation ability with advanced density based operation when intelligent systems including neural networks and fuzzy inference systems can be generated by clustering results. Also, results of the mechanism will be served to information of decision support system to infer the useful knowledge. It can extend to healthcare and medical decision support system to help experts or specialists. We show and explain the usefulness of the proposed method using simulation and results.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.40 no.2
• /
• pp.92-98
• /
• 2017

Data clustering determines a group of patterns using similarity measure in a dataset and is one of the most important and difficult technique in data mining. Clustering can be formally considered as a particular kind of NP-hard grouping problem. K-means algorithm which is popular and efficient, is sensitive for initialization and has the possibility to be stuck in local optimum because of hill climbing clustering method. This method is also not computationally feasible in practice, especially for large datasets and large number of clusters. Therefore, we need a robust and efficient clustering algorithm to find the global optimum (not local optimum) especially when much data is collected from many IoT (Internet of Things) devices in these days. The objective of this paper is to propose new Hybrid Simulated Annealing (HSA) which is combined simulated annealing with K-means for non-hierarchical clustering of big data. Simulated annealing (SA) is useful for diversified search in large search space and K-means is useful for converged search in predetermined search space. Our proposed method can balance the intensification and diversification to find the global optimal solution in big data clustering. The performance of HSA is validated using Iris, Wine, Glass, and Vowel UCI machine learning repository datasets comparing to previous studies by experiment and analysis. Our proposed KSAK (K-means+SA+K-means) and SAK (SA+K-means) are better than KSA(K-means+SA), SA, and K-means in our simulations. Our method has significantly improved accuracy and efficiency to find the global optimal data clustering solution for complex, real time, and costly data mining process.

• Proceedings of the IEEK Conference
• /
• 2003.11c
• /
• pp.135-139
• /
• 2003

Sensor networks consist of sensor nodes with small-size, low-cost, low-power, and multi-functions to sense, to process and to communicate. Minimizing power consumption of sensors is an important issue in sensor networks due to limited power in sensor networks. Clustering is an efficient way to reduce data flow in sensor networks and to maintain less routing information. In this paper, we propose a multi-hop clustering mechanism using global and local ID to reduce transmission power consumption and an efficient routing method for improved data fusion and transmission.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.65 no.9
• /
• pp.1541-1550
• /
• 2016

In this paper, we introduce a novel learning methodology of fuzzy clustering-based neural network pattern classifier. Fuzzy clustering-based neural network pattern classifier depicts the patterns of given classes using fuzzy rules and categorizes the patterns on unseen data through fuzzy rules. Least squares estimator(LSE) or weighted least squares estimator(WLSE) is typically used in order to estimate the coefficients of polynomial function, but this study proposes a novel coefficient estimate method which includes advantages of the existing methods. The premise part of fuzzy rule depicts input space as "If" clause of fuzzy rule through fuzzy c-means(FCM) clustering, while the consequent part of fuzzy rule denotes output space through polynomial function such as linear, quadratic and their coefficients are estimated by the proposed local least squares estimator(LLSE)-based learning. In order to evaluate the performance of the proposed pattern classifier, the variety of machine learning data sets are exploited in experiments and through the comparative analysis of performance, it provides that the proposed LLSE-based learning method is preferable when compared with the other learning methods conventionally used in previous literature.