Comparative Analysis between K-Means and K-Medoids for Statistical Clustering

Arbin, Norazam; Suhaimi, Nur Suhailayani; Mokhtar, Nurul Zafirah; Othman, Zalinda

doi:10.1109/aims.2015.82

Cited by 32 publications

(21 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Moreover, a k-means algorithm has the excellent ability to handle a large number of investigated data. The k-means algorithm applies a standard distance measure formula, to determine the similarity of the data repetitively, to obtain the high inter-cluster distance among clusters (Arbin et al, 2016). K-means clustering iteratively finds the k centroids and assigns every object to the nearest centroid (Park & Jun, 2009).…”

Section: Clusteringmentioning

confidence: 99%

Measuring Performance by Integrating K-Medoids With Dea: Mongolian Case

Bayaraa

Tarnóczi

Fenyves

2019

Journal of Business Economics and Management

View full text Add to dashboard Cite

Performance measurement encourages Decision Making Units (DMUs) to improve their level of performance by comparing their current financial positions with that of their peers. Data Envelopment Analysis (DEA) is a widely used approach to performance measurement, though it is susceptible when the data is heterogeneous. The main objective of this study is to examine the performance of Mongolian listed companies by combining DEA and a k-medoid clustering method. Clustering facilitates the characterization and patterns of data and identification of homogenous groups. This study applies the integration of k-medoids and performance measurement. The research used 89 Mongolian companies' financial statements from 2012 to 2015 -obtained from the Mongolian Stock Exchange website. The companies are grouped by k-medoids clustering, and efficiency of each cluster is evaluated by DEA. According to the silhouette method, the companies are classified into two clusters which are considered first cluster as small and medium-sized (80), and second cluster as big (9) companies. Both clusters are analyzed and compared by financial ratios. The mean efficiency score of big companies' is much higher than that of small and medium-sized companies. Integrated results show that cluster-specific efficiency provides better performance than pre-clustering efficiency results.Keywords: financial performance, k-medoids clustering, data envelopment analysis, input efficiency, variable return to scale, decision making unit. JEL Classification: C38, C14, L25.ing (finding meaningful groups of objects that share common characteristics) and utility (to abstract the representative object from among many others in the same clusters) (Wu, 2012).Clustering techniques are divided into partitional and hierarchical types. The most popular and well-known partitional cluster technique is k-means, which is widely employed in research. Although k-means is a popular choice among partitional clusters, it is sensitive to outliers. On the contrary, the k-medoids algorithm is more robust and less sensitive to outliers. Research, which compared k-medoids with k-means, suggested the k-medoid was better in all aspects. For example, Arora and Varshney (2016) compared k-means and k-medoids in their research. Their results proved that k-medoids is better than k-means; as execution time, sensitivity to outliers and space complexity of overlapping are all less. Narayana and Vasumathi (2018) stated in their work that the k-medoids technique is more accurate and easier to understand than k-means clustering. Moreover, Patel and Singh (2013) studied a new approach for k-means and k-medoids algorithm and concluded that k-medoids improved accuracy. However, Arbin, Suhaimi, Mokhtar, and Othman (2016) evaluated k-means and k-medoids, and both methods were found to be good having mean errors less than three.The K-medoids algorithm, which was proposed by Kaufman and Rousseeuw (1987), was developed and investigated by various researchers from different fields. For example, Ho-Kieu, Vo-Van...

show abstract

Section: Clusteringmentioning

confidence: 99%

Measuring Performance by Integrating K-Medoids With Dea: Mongolian Case

Bayaraa

Tarnóczi

Fenyves

2019

Journal of Business Economics and Management

View full text Add to dashboard Cite

show abstract

“…According to [8], understanding the use of appropriate approaches in data analysis becomes specific to get the expected results. K-Means analysis provides good performance records in comparative analysis between K-Means and K-Medoids for Statistical Clustering.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Many studies have been conducted by researchers related to the use of k-means algorithms in grouping. K-Means clustering is one of the most widely used grouping techniques in research [6]- [8], [10]- [15]. K-means shows more accurate results for solid datasets compared to rare datasets.…”

Section: F K-means Clusteringmentioning

confidence: 99%

Determining Feasibility Level of Beef Quality Based on Histogram and K-Means Clustering

Winiarti

Azhari

Agusta

2018

2018 International Symposium on Advanced Intelligent Informatics (SAIN)

View full text Add to dashboard Cite

Beef is one of the staple foods that are consumed by many people. The rise of beef with poor quality on the market and the lack of knowledge of consumers and the public in distinguishing the quality of meat that is decent and not suitable for consumption when seen in plain view. The Department of Agriculture and Food of the City of Yogyakarta, especially Animal Slaughterhouse (RPH), Giwangan Yogyakarta in determining the quality of beef, still uses laboratory tests which take a long time. This study proposes a system to identify the image of beef that is suitable for consumption based on the color features generated from the image histogram with Hue, Saturation, and Intensity (HSI) color models and the K-Means Clustering method. The results of identification carried out on 40 test data obtained maximum and minimum values of HSI from each feasibility category.

show abstract

“…In recent years, many researchers have proposed density-based data stream clustering algorithms. However, several issues related to these clustering algorithms must be considered [13,[52][53][54][55][56][57][58][59], such as most are not entirely online methods, are unable to handle evolving data streams, are unable to manage the noisy characteristics of data streams, or suffer from high memory requirements, low processing rates, or the "curse of dimensionality" [45,[60][61][62][63]. Moreover, the existing density-based clustering algorithms have high computational times and low cluster quality for clustering data streams.…”

Section: Introductionmentioning

confidence: 99%

Online Clustering of Evolving Data Streams Using a Density Grid-Based Method

et al. 2020

View full text Add to dashboard Cite

In recent years, a significant boost in data availability for persistent data streams has been observed. These data streams are continually evolving, with the clusters frequently forming arbitrary shapes instead of regular shapes in the data space. This characteristic leads to an exponential increase in the processing time of traditional clustering algorithms for data streams. In this study, we propose a new online method, which is a density grid-based method for data stream clustering. The primary objectives of the density grid-based method are to reduce the number of distant function calls and to improve the cluster quality. The method is conducted entirely online and consists of two main phases. The first phase generates the Core Micro-Clusters (CMCs), and the second phase combines the CMCs into macro clusters. The gridbased method was utilized as an outlier buffer in order to handle multi-density data and noises. The method was tested on real and synthetic data streams employing different quality metrics and was compared with the popular method of clustering evolving data streams into arbitrary shapes. The proposed method was demonstrated to be an effective solution for reducing the number of calls to the distance function and improving the cluster quality. INDEX TERMS Clustering, data stream, evolving, grid-based method, core-micro-cluster, online.

show abstract

Comparative Analysis between K-Means and K-Medoids for Statistical Clustering

Cited by 32 publications

References 6 publications

Measuring Performance by Integrating K-Medoids With Dea: Mongolian Case

Measuring Performance by Integrating K-Medoids With Dea: Mongolian Case

Determining Feasibility Level of Beef Quality Based on Histogram and K-Means Clustering

Online Clustering of Evolving Data Streams Using a Density Grid-Based Method

Contact Info

Product

Resources

About