Critical analysis of DBSCAN variations

Ali, Tariq; Asghar, Sohail; Sajid, Naseer Ahmed

doi:10.1109/iciet.2010.5625720

Cited by 53 publications

(27 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to DBSCAN's popularity among density-based clustering algorithms, optimization and parallelization of the algorithm has been widely studied [5]. We first explain the DBSCAN algorithm in detail, then present previous parallelization efforts that are most significant to the parallelization style of Mr. Scan along with the most scalable algorithms.…”

Section: Background and Related Workmentioning

confidence: 99%

Mr. Scan

Welton

Samanas

Miller

2013

Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis

View full text Add to dashboard Cite

Density-based clustering algorithms are a widely-used class of data mining techniques that can find irregularly shaped clusters and cluster data without prior knowledge of the number of clusters it contains. DBSCAN is the most wellknown density-based clustering algorithm. We introduce our version of DBSCAN, called Mr. Scan, which uses a hybrid parallel implementation that combines the MRNet treebased distribution network with GPGPU-equipped nodes. Mr. Scan avoids the problems of existing implementations by effectively partitioning the point space and by optimizing DBSCAN's computation over dense data regions. We tested Mr. Scan on both a geolocated Twitter dataset and image data obtained from the Sloan Digital Sky Survey. At its largest scale, Mr. Scan clustered 6.5 billion points from the Twitter dataset on 8,192 GPU nodes on Cray Titan in 17.3 minutes. All other parallel DBSCAN implementations have only demonstrated the ability to cluster up to 100 million points.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

Mr. Scan

Welton

Samanas

Miller

2013

Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis

View full text Add to dashboard Cite

show abstract

“…Evaluation of attributes with class variable is as shown below: (1). Experimental evaluation using Dbscan Algorithm: Dbscan algorithm makes clusters by iteratively [4] checking neighbor elements of each data points within dataset [11]. In case nearby elements are more than minPts, a new cluster formed with O as core object.…”

Section: Expermimantal Evaluationmentioning

confidence: 99%

Clustered Comparative Analysis of Security Sensor Discrimination Data

Bhargava

Jain²,

Kumar³

et al. 2018

Annals of Computer Science and Information Systems

View full text Add to dashboard Cite

Abstract-Security alarm is used to protect from burglary (theft), property damage and from intruders. These security alarms consists sensors and alerting device to indicate the intrusion. Clustering is data mining technique which is used to analyzing the data. In this paper we discus about different clustering algorithm like DBSCAN, Farthest first. These algorithms are used to evaluate the different number of clusters with the sensor discrimination data base. In any organization Sensor security has many types of security alarm. It may be glass breaking alarm, smoke heat and carbon monoxide alarm, and it may be false alarm. Our aim is to compare the different algorithms with the sensors data to find density clusters i.e. which type of data will provide dense cluster of useful alarm condition. This evaluation will also detect the outliers within data such as empty alarms.

show abstract

“…Fuzzy c-means, proposed by Bezdek [17], similar to k-means but using fuzzy logic, where every point belongs to every cluster with some degree; quality-threshold clustering proposed by Heyer, Kruglyak and Yooseph in 1999 [18], designed for gene clustering and requires only maximum diameter for clusters; agglomerative and divisive hierarchical clustering [19]; and many others. More information about recent developments in DBSCAN clustering method can be found in [20] III. FAULT DETECTION METHOD Values we are referring to as partially regular are such values as are presented in Fig.…”

Section: Cluster Analysis and Algorithmsmentioning

confidence: 99%

Automated fault detection method in process data based on cluster analysis

Belić¹,

Hocenski

2014

2014 IEEE 23rd International Symposium on Industrial Electronics (ISIE)

View full text Add to dashboard Cite

This article presents a method for detecting changes in behavior of data. It is based on cluster analysis, which is a common name for methods that group data in segments called clusters, based on similarities and differences of data itself, without supervision of human observer. The data analyzed by clustering techniques are commonly met in process industry: locally constant process values with a lot of noise and sudden changes to completely different values. The experimental application was developed for evaluation of proposed method and gained results prove its quality for several data patterns. This method can be used for automated fault detection applied to industrial process data when data errors are more complex than simple breaching of data limits or minimum and maximum.

show abstract

Critical analysis of DBSCAN variations

Cited by 53 publications

References 19 publications

Mr. Scan

Mr. Scan

Clustered Comparative Analysis of Security Sensor Discrimination Data

Automated fault detection method in process data based on cluster analysis

Contact Info

Product

Resources

About