Hybrid framework for DBSCAN algorithm using fuzzy logic

Author

Beri, Saefia ; Kaur, Kamaljit

Author_Institution

Dept. of Comput. Sci. & Eng., Guru Nanak Dev Univ., Amritsar, India

fYear

2015

fDate

25-27 Feb. 2015

Firstpage

383

Lastpage

387

Abstract

Data mining process is to obtain information from a data set and then convert it into an understandable and meaningful information for further use. DBSCAN, a density based clustering algorithm, identifies clusters of varying shape and outliers. DBSCAN is based on bivalent logic. Therefore it can only detect objects as completely belonging to a particular cluster or not wholly belonging to it. In this paper, a framework of methodology of DBSCAN algorithm with the integration of fuzzy logic is proposed. The extent to which an object belongs to a particular cluster will be determined using membership values. The improved version of DBSCAN algorithm will be the hybridization of DBSCAN algorithm with fuzzy if-then rules.

Keywords

data mining; fuzzy logic; fuzzy reasoning; pattern clustering; DBSCAN algorithm; bivalent logic; data mining process; density based clustering algorithm; density-based spatial clustering-of-application-with-noise; fuzzy if-then rules; fuzzy logic; hybrid framework; membership values; Algorithm design and analysis; Breast cancer; Classification algorithms; Clustering algorithms; Data mining; Noise; Spatial databases; DBSCAN; bivalent logic; clustering;

fLanguage

English

Publisher

ieee

Conference_Titel

Futuristic Trends on Computational Analysis and Knowledge Management (ABLAZE), 2015 International Conference on

Conference_Location

Noida

Print_ISBN

978-1-4799-8432-9

Type

conf

DOI

10.1109/ABLAZE.2015.7155024

Filename

7155024