A Vibration Method for Discovering Density Varied Clusters

Elbatta, Mohammad T.; Bolbol, Raed M.; Ashour, Wesam M.

doi:10.5402/2012/723516

Cited by 22 publications

(19 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Then it "vibrates" points towards the cluster that has the maximum effect on these points. DMDBSCAN (Dynamic Method DBSCAN) [14] is a new enhancement of DBSCAN which has pointed out that in clusters, generated by DBSCAN, there is wide density variation. Compared to DBSCAN which uses global Eps.…”

Section: Related Workmentioning

confidence: 99%

AMF-IDBSCAN: Incremental Density Based Clustering Algorithm Using Adaptive Median Filtering Technique

Chefrour¹,

Souici-Meslati²

2019

IJCAI

View full text Add to dashboard Cite

Density-based spatial clustering of applications with noise (DBSCAN) is a fundamental algorithm for density-based clustering. It can discover clusters of arbitrary shapes and sizes from a large amount of data, which contains noise and outliers. However, it fails to treat large datasets, outperform when new objects are inserted into the existing database, remove noise points or outliers totally and handle the local density variation that exists within the cluster. So, a good clustering method should allow a significant density modification within the cluster and should learn dynamics and large databases. In this paper, an enhancement of the DBSCAN algorithm is proposed based on incremental clustering called AMF-IDBSCAN which builds incrementally the clusters of different shapes and sizes in large datasets and eliminates the presence of noise and outliers. The proposed AMF-IDBSCAN algorithm uses a canopy clustering algorithm for pre-clustering the data sets to decrease the volume of data, applies an incremental DBSCAN for clustering the data points and Adaptive Median Filtering (AMF) technique for post-clustering to reduce the number of outliers by replacing noises by chosen medians. Experiments with AMF-IDBSCAN are performed on the University of California Irvine (UCI) repository UCI data sets. The results show that our algorithm performs better than DBSCAN, IDBSCAN, and DMDBSCAN. Povzetek: V članku je predstavljen nov algoritem AMF-IDBSCAN, izboljšana različica DBSCAN, ki uporablja grozdenje krošenj za zmanjšanje obsega podatkov in tehnike AMF za odpravo hrupa.

show abstract

Section: Related Workmentioning

confidence: 99%

AMF-IDBSCAN: Incremental Density Based Clustering Algorithm Using Adaptive Median Filtering Technique

Chefrour¹,

Souici-Meslati²

2019

IJCAI

View full text Add to dashboard Cite

show abstract

“…Keduanya sangat efektif untuk proses clustering dengan jumlah data yang cukup besar. Sedang Elbatta (Elbatta 2013) melakukan penelitian dengan menggunakan density threshold untuk menentukan variasi nilai ε pada proses clustering, dimana masih harus menentukan threshold yang sesuai untuk mendapatkan nilai ε yang optimal.…”

Section: Introductionunclassified

Clustering the Potential Risk of Tsunami Using Density-Based Spatial Clustering of Application With Noise (Dbscan)

Furqon

Muflikhah

2016

JEEST

View full text Add to dashboard Cite

Tsunami is one of the deadliest natural disaster that causing devastating property damage and loss of life. Therefore, this triggers many scientist to do researches in tsunami mitigation disaster, such as analyzing the potential risks caused by tsunami. The process of analyzing the potential risk caused by tsunami can be done by grouping the data of tsunami based on characteristics of the previous tsunami events. DBSCAN (Density-based Spatial Clustering of Application with Noise) is a popular clustering method and can be used to do the task. The algorithm do the clustering processes using density-based concept that able to detect outlier/noise and clusters irregular shapes. It was proved in this research where the evaluation method using Silhouette Coefficient on the DBSCAN clustering result gave highest value 0.96056649 for ε and minPts value of .1 and 0.1.

show abstract

“…DMDBSCAN [21] algorithm is an extension of DBSCAN, it depends on selecting several values for Eps from the k-dist plot of data based on seeing sharp change on this curve, but this method does not work well with the presence of noise, and sometimes lead to split some clusters.…”

Section: Related Workmentioning

confidence: 99%

A Clustering Algorithm based on Local Density of Points

Fahim¹

2017

IJMECS

View full text Add to dashboard Cite

Abstract-Data clustering is very active and attractive research area in data mining; there are dozens of clustering algorithms that have been published. Any clustering algorithm aims to classify data points according to some criteria. DBSCAN is the most famous and well-studied algorithm. Clusters are recorded as dense regions separated from each other by spars regions. It is based on enumerating the points in Epsneighborhood of each point. This paper proposes a clustering method based on k-nearest neighbors and local density of objects in data; that is computed as the total of distances to the most near points that affected on it. Cluster is defined as a continuous region that has points within local densities fall between minimum local density and maximum local density. The proposed method identifies clusters of different shapes, sizes, and densities. It requires only three parameters; these parameters take only integer values. So it is easy to determine. The experimental results demonstrate the superior of the proposed method in identifying varied density clusters.

show abstract

A Vibration Method for Discovering Density Varied Clusters

Cited by 22 publications

References 19 publications

AMF-IDBSCAN: Incremental Density Based Clustering Algorithm Using Adaptive Median Filtering Technique

AMF-IDBSCAN: Incremental Density Based Clustering Algorithm Using Adaptive Median Filtering Technique

Clustering the Potential Risk of Tsunami Using Density-Based Spatial Clustering of Application With Noise (Dbscan)

A Clustering Algorithm based on Local Density of Points

Contact Info

Product

Resources

About