Hybrid Clustering Method Research Articles

KC-Means: A Fast Fuzzy Clustering

A novel hybrid clustering method, named KC-Means clustering, is proposed for improving upon the clustering time of the Fuzzy C-Means algorithm. The proposed method combines K-Means and Fuzzy C-Means algorithms into two stages. In the first stage, the K-Means algorithm is applied to the dataset to find the centers of a fixed number of groups. In the second stage, the Fuzzy C-Means algorithm is applied on the centers obtained in the first stage. Comparisons are then made between the proposed and other algorithms in terms of time processing and accuracy. In addition, the mentioned clustering algorithms are applied to a few benchmark datasets in order to verify their performances. Finally, a class of Minkowski distances is used to determine the influence of distance on the clustering performance.

Application of hybrid clustering methods for student performance evaluation

In the present research paper, an attempt has been made to design for evaluating the assessment of student in the educational environment by using the hybrid clustering methods. Various issues along problems are associated with the academic performance evaluation in the field of education. The proposed new approach known as hybrid clustering is based on integrated techniques of Subtractive and Fuzzy C-Means clustering methods. The assessment of student academic performance can be considered as a clustering problem. The clusters are formed on the basis of the intelligence level of students. The intelligence level wise grouping is essential for maintaining the homogeneity of the group. Otherwise it would be difficult to provide good educational services to the highly diverse student population.

An efficient hybrid clustering method based on improved cuckoo optimization and modified particle swarm optimization algorithms

Partitional data clustering with K-means algorithm is the dividing of objects into smaller and disjoint groups that has the most similarity with objects in a group and most dissimilarity from the objects of other groups. Several techniques have been proposed to avoid the major limitations of K-Means such as sensitive to initialization and easily convergence to local optima. An alternative to solve the drawback of the sensitive to centroids’ initialization in K-Means is the K-Harmonic Means (KHM) clustering algorithm. However, KHM is sensitive to the noise and easily runs into local optima. Over the past decade, many algorithms are developed for solving this problems based on evolutionary method. However, each algorithm has its own advantages, limitations and shortcomings. In this paper, we combined K-Harmonic Means (KHM) clustering algorithm with an improved Cuckoo Search (ICS) and particle swarm optimization (PSO). ICS is intended to global optimum solution using Lévy flight method through changing radius in a dynamic and shrewd manner. Therefore, it is faster than standard cuckoo search. ICS is effected with PSO to avoid falling into local optima. The proposed algorithm, called ICMPKHM, solves the local optima problem of KHM with significant improvement on efficacy and stability. Experiments with benchmark datasets show that the proposed algorithm is quite insensitive to the centroids’ initialization. Comparative studies with other algorithms reveal that the proposed algorithm produce high quality and stable clustering results.

A hybrid clustering and ranking method for best positioned logistics distribution centre in Balkan Peninsula

AbstractFacility location decisions are critical in strategic planning for a wide range of operational and logistical decisions. Facility location problem with focus on logistics distribution centre (LDC) in Balkan Peninsula (BP) is discussed in this article. Methodological hybrid genetic algorithm, Analytical Hierarchy Process, and fuzzy c-means method is proposed here and it is shown how such a model can be of assistance in analysing a multi criteria decision-making problem. This research represents continuation of three existing studies. The experimental results in our research could be well compared with other official results of the feasibility study of the LDC located in BP.

SPATIO-TEMPORAL CHARACTERISTICS OF RESIDENT TRIP BASED ON POI AND OD DATA OF FLOAT CAR IN BEIJING

Abstract. Due to the influence of the urban inherent regional functional distribution, the daily activities of the residents presented some spatio-temporal patterns (periodic patterns, gathering patterns, etc.). In order to further understand the spatial and temporal characteristics of urban residents, this paper research takes the taxi trajectory data of Beijing as a sample data and studies the spatio-temporal characteristics of the residents' activities on the weekdays. At first, according to the characteristics of the taxi trajectory data distributed along the road network, it takes the Voronoi generated by the road nodes as the research unit. This paper proposes a hybrid clustering method – based on grid density, which is used to cluster the OD (origin and destination) data of taxi at different times. Then，combining with the POI data of Beijing, this research calculated the density of the POI data in the clustering results, and analyzed the relationship between the activities of residents in different periods and the functional types of the region. The final results showed that the residents were mainly commuting on weekdays. And it found that the distribution of travel density showed a concentric circle of the characteristics, focusing on residential areas and work areas. The results of cluster analysis and POI analysis showed that the residents' travel had experienced the process of "spatial relative dispersion – spatial aggregation – spatial relative dispersion" in one day.

Short-term wind power forecasts by a synthetical similar time series data mining method

As the aggravating influence of growing wind power, wind power forecasting research becomes more important in economic operation and safety management of power system. A novel short-term wind power forecasting methodology consists of a hybrid clustering method and a wavelet based neural network is introduced. The clustering similar measure function combines the Euclidean Distance and Angle Cosine together, aims to identify the similar wind speed days which are close in space distance and have similar variance trend synthetically. Then similar daily samples as the predicting days are treated as training samples of an improved particle swarm optimization based wavelet neural network. The proposed forecasting strategy is applied to two real wind farms in China. The results demonstrate that the strategy can identify the similar time series and improve the predicting accuracy effectively, compared with some other forecasting models.

A Hybridized Clustering Approach based on Rough Set and Fuzzy c-Means to Mine Cholesterol Sequence from ABC Family

Objectives: The current study is focused on design of a computational model for human ABC transporters; wherein the TM-sequences matching the CRAC/CARC motif are extracted. Methods: The postulation of cholesterol binding motif (CRAC/CARC), its presence in different proteins and validating its interaction with cholesterol has indeed established the importance of the motif in cholesterol-mediated modulation of protein/signaling pathway. Several viral proteins and membrane proteins (especially alpha-helical trans membrane proteins) such as GPCR transporters are reported to be modulated by cholesterol. The experimental studies are so far performed on only a few proteins in a family but based on an evolutionary conservation and consensus an exploration can be done confidently within a family. However, the representation of motif has a low consensus yielding several false positives thus reducing its reliability. Findings: A computational hybrid clustering method based on rough set with fuzzy c-means algorithm is used to mine the cholesterol sequence from ABC family. Higher weightage is given to those sequences based on the following parameters: motifs with more number of sub motifs, number of helices bearing the motif in a protein and compliance with the orientation of the cholesterol in the membrane for its interaction with the motif. Improvement: A detailed study in a given super family with an approach to reduce redundancy and enrichment can improve its predictability.

A Hybrid Clustering Method Using Balanced Scorecard and Data Envelopment Analysis

This paper introduces a new hybrid clustering method using Data Envelopment Analysis (DEA) and Balanced Scorecard (BSC) methods. DEA cannot identify its’ input and output itself, and it is a major weakness of the DEA. In the proposed method, this gap is resolved by integrating DEA with BSC. Some decision-making units (DMUs) needed in DEA method, in compliance with some inputs and outputs is the major drawback of this integration. To deal with this disadvantage, the proposed method selects the most important strategic factors, attained from the BSC method. These data considered to be the input data for the DEA method to calculate relative closeness (RC) of each DMU to the ideal one. Plotting the screen diagram regarding RC index leads us to the final clustering method. Finally, computational results show the applicability and usefulness of the method.

A Hybrid Clustering Method Using Balanced Scorecard and Data Envelopment Analysis

This paper introduces a new hybrid clustering method using Data Envelopment Analysis (DEA) and Balanced Scorecard (BSC) methods. DEA cannot identify its’ input and output itself, and it is a major weakness of the DEA. In the proposed method, this gap is resolved by integrating DEA with BSC. Some decision-making units (DMUs) needed in DEA method, in compliance with some inputs and outputs is the major drawback of this integration. To deal with this disadvantage, the proposed method selects the most important strategic factors, attained from the BSC method. These data considered to be the input data for the DEA method to calculate relative closeness (RC) of each DMU to the ideal one. Plotting the screen diagram regarding RC index leads us to the final clustering method. Finally, computational results show the applicability and usefulness of the method.

D-FICCA: A density-based fuzzy imperialist competitive clustering algorithm for intrusion detection in wireless sensor networks

Owing to the scattered nature of Denial-of-Service attacks, it is tremendously challenging to detect such malicious behavior using traditional intrusion detection systems in Wireless Sensor Networks (WSNs). In the current paper, a hybrid clustering method is introduced, namely a density-based fuzzy imperialist competitive clustering algorithm (D-FICCA). Hereby, the imperialist competitive algorithm (ICA) is modified with a density-based algorithm and fuzzy logic for optimum clusteringin WSNs. A density-based clustering algorithm helps improve the imperialist competitive algorithm for the formation of arbitrary cluster shapes as well as handling noise. The fuzzy logic controller (FLC) assimilates to imperialistic competition by adjusting the fuzzy rules to avoid possible errors of the worst imperialist action selection strategy. The proposed method aims to enhance the accuracy of malicious detection. D-FICCA is evaluated on a publicly available dataset consisting of real measurements collected from sensors deployed at the Intel Berkeley Research Lab. Its performance is compared against existing empirical methods, such as K-MICA, K-mean, and DBSCAN. The results demonstrate that the proposed framework achieves higher detection accuracy 87% and clustering quality 0.99 compared to existing approaches.

Choosing DBSCAN Parameters Automatically using Differential Evolution

Over the last several years, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) has been widely applied in many areas of science due to its simplicity, robustness against noise (outlier) and ability to discover clusters of arbitrary shapes. However, DBSCAN algorithm requires two initial input parameters, namely Eps (the radius of the cluster) and MinPts (the minimum data objects required inside the cluster) which both have a significant influence on the clustering results. Hence, DBSCAN is sensitive to its input parameters and it is hard to determine them a priori. This paper presents an efficient and effective hybrid clustering method, named BDE-DBSCAN, that combines Binary Differential Evolution and DBSCAN algorithm to simultaneously quickly and automatically specify appropriate parameter values for Eps and MinPts. Since the Eps parameter can largely degrades the efficiency of the DBSCAN algorithm, the combination of an analytical way for estimating Eps and Tournament Selection (TS) method is also employed. Experimental results indicate the proposed method is precise in determining appropriate input parameters of DBSCAN algorithm.

Detecting Image Forgery Based on Noise Estimation

With the advent of the Internet and low-price digital cameras, as well as powerful image editing software, the authenticity of digital images can no longer be taken for granted. Image noises are often introduced into the tampered region during image manipulation process. In this paper, we propose a detection method to locate image forgeries based on noise estimation on HSV color space and hybrid clustering method combined with unsupervised clustering and supervised clustering. A suspicious image is first converted into HSV color space and segmented into non-overlapping image blocks. Then the noise variance at each local image block is estimated as input of unsupervised clustering. Finally, a supervised clustering method based on SVM is used to further improve the detection accuracy. Our experimental results demonstrate that the proposed method can effectively expose tampered regions from tampered images.

Mining categorical sequences from data using a hybrid clustering method

The identification of different dynamics in sequential data has become an every day need in scientific fields such as marketing, bioinformatics, finance, or social sciences. Contrary to cross-sectional or static data, this type of observations (also known as stream data, temporal data, longitudinal data or repeated measures) are more challenging as one has to incorporate data dependency in the clustering process. In this research we focus on clustering categorical sequences. The method proposed here combines model-based and heuristic clustering. In the first step, the categorical sequences are transformed by an extension of the hidden Markov model into a probabilistic space, where a symmetric Kullback–Leibler distance can operate. Then, in the second step, using hierarchical clustering on the matrix of distances, the sequences can be clustered. This paper illustrates the enormous potential of this type of hybrid approach using a synthetic data set as well as the well-known Microsoft dataset with website users search patterns and a survey on job career dynamics.

OLYMPUS: An automated hybrid clustering method in time series gene expression. Case study: Host response after Influenza A (H1N1) infection

The increasing flow of short time series microarray experiments for the study of dynamic cellular processes poses the need for efficient clustering tools. These tools must deal with three primary issues: first, to consider the multi-functionality of genes; second, to evaluate the similarity of the relative change of amplitude in the time domain rather than the absolute values; third, to cope with the constraints of conventional clustering algorithms such as the assignment of the appropriate cluster number. To address these, we propose OLYMPUS, a novel unsupervised clustering algorithm that integrates Differential Evolution (DE) method into Fuzzy Short Time Series (FSTS) algorithm with the scope to utilize efficiently the information of population of the first and enhance the performance of the latter. Our hybrid approach provides sets of genes that enable the deciphering of distinct phases in dynamic cellular processes.We proved the efficiency of OLYMPUS on synthetic as well as on experimental data. The discriminative power of OLYMPUS provided clusters, which refined the so far perspective of the dynamics of host response mechanisms to Influenza A (H1N1). Our kinetic model sets a timeline for several pathways and cell populations, implicated to participate in host response; yet no timeline was assigned to them (e.g. cell cycle, homeostasis). Regarding the activity of B cells, our approach revealed that some antibody-related mechanisms remain activated until day 60 post infection.The Matlab codes for implementing OLYMPUS, as well as example datasets, are freely accessible via the Web (http://biosignal.med.upatras.gr/wordpress/biosignal/).

A powerful hybrid clustering method based on modified stem cells and Fuzzy C-means algorithms

One of the simple techniques for Data Clustering is based on Fuzzy C-means (FCM) clustering which describes the belongingness of each data to a cluster by a fuzzy membership function instead of a crisp value. However, the results of fuzzy clustering depend highly on the initial state selection and there is also a high risk for getting the best results when the datasets are large. In this paper, we present a hybrid algorithm based on FCM and modified stem cells algorithms, we called it SC-FCM algorithm, for optimum clustering of a dataset into K clusters. The experimental results obtained by using the new algorithm on different well-known datasets compared with those obtained by K-means algorithm, FCM, Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Artificial Bee Colony (ABC) Algorithm demonstrate the better performance of the new algorithm.

New method for extracting region of interest of dynamic PET images based on curve clustering

Concerning the problem that many current clustering methods based on kinetic characteristics ignore the continuous temporal information of Time Activity Curve(TAC),a method for Region Of Interest(ROI) extraction based on curve clustering was proposed in the paper.The proposed method contains three steps.Firstly,K-Means algorithm was used to remove the background to obtain a coarse mask of the heart.Secondly,curve clustering was used to extract myocardium from the heart obtained in the first step.Finally,blood cavity was delineated based on spatial relationship between pixels.The method was applied to extract the ROI from fourteen mouse PET images.The experimental results indicate that the proposed method is more accurate in delineating blood cavity of the fourteen mice than K-Means and Hybrid Clustering Method(HCM),and it is more precise and stable.

Clustering methods applied in the detection of Ki67 hot‐spots in whole tumor slide images: An efficient way to characterize heterogeneous tissue‐based biomarkers

Whole-slide scanners allow the digitization of an entire histological slide at very high resolution. This new acquisition technique opens a wide range of possibilities for addressing challenging image analysis problems, including the identification of tissue-based biomarkers. In this study, we use whole-slide scanner technology for imaging the proliferating activity patterns in tumor slides based on Ki67 immunohistochemistry. Faced with large images, pathologists require tools that can help them identify tumor regions that exhibit high proliferating activity, called "hot-spots" (HSs). Pathologists need tools that can quantitatively characterize these HS patterns. To respond to this clinical need, the present study investigates various clustering methods with the aim of identifying Ki67 HSs in whole tumor slide images. This task requires a method capable of identifying an unknown number of clusters, which may be highly variable in terms of shape, size, and density. We developed a hybrid clustering method, referred to as Seedlink. Compared to manual HS selections by three pathologists, we show that Seedlink provides an efficient way of detecting Ki67 HSs and improves the agreement among pathologists when identifying HSs.

Comparing exposure metrics for classifying ‘dangerous heat’ in heat wave and health warning systems

Heat waves have been linked to excess mortality and morbidity, and are projected to increase in frequency and intensity with a warming climate. This study compares exposure metrics to trigger heat wave and health warning systems (HHWS), and introduces a novel multi-level hybrid clustering method to identify potential dangerously hot days. Two-level and three-level hybrid clustering analysis as well as common indices used to trigger HHWS, including spatial synoptic classification (SSC), and the 90th, 95th, and 99th percentiles of minimum and relative minimum temperature (using a 10day reference period), were calculated using a summertime weather dataset in Detroit from 1976 to 2006. The days classified as ‘hot’ with hybrid clustering analysis, SSC, minimum and relative minimum temperature methods differed by method type. SSC tended to include the days with, on average, 2.5°C lower daily minimum temperature and 5.3°C lower dew point than days identified by other methods. These metrics were evaluated by comparing their performance in predicting excess daily mortality. The 99th percentile of minimum temperature was generally the most predictive, followed by the three-level hybrid clustering method, the 95th percentile of minimum temperature, SSC and others. Our proposed clustering framework has more flexibility and requires less substantial meteorological prior information than the synoptic classification methods. Comparison of these metrics in predicting excess daily mortality suggests that metrics thought to better characterize physiological heat stress by considering several weather conditions simultaneously may not be the same metrics that are better at predicting heat-related mortality, which has significant implications in HHWSs.

Using hybrid data mining and machine learning clustering analysis to predict the turnover rate for technology professionals

This study applies clustering analysis for data mining and machine learning to predict trends in technology professional turnover rates, including the hybrid artificial neural network and clustering analysis known as the self-organizing map (SOM). This hybrid clustering method was used to study the individual characteristics of turnover trend clusters. Using a transaction questionnaire, we studied the period of peak turnover, which occurs after the Chinese New Year, for individuals divided into various age groups. The turnover trend of technology professionals was examined in well-known Taiwanese companies. The results indicate that the high outstanding turnover trend circle was primarily caused by a lack of inner fidelity identification, leadership and management. Based on cross-verification, the clustering accuracy rate was 92.7%. This study addressed problems related to the rapid loss of key human resources and should help organizations learn how to enhance competitiveness and efficiency.

Optimal and hierarchical clustering of large-scale hybrid networks for scientific mapping

Previous studies have shown that hybrid clustering methods based on textual and citation information outperforms clustering methods that use only one of these components. However, former methods fo...