Evaluating and comparing between fuzzy K-mean and K-mean clustering in general cloud

Other Title(s)

تقييم و مقارنة خوارزمية تجميع البيانات الضبابية الوسطية و خوارزمية تجميع البيانات الوسطية في البيئة السحابية

Dissertant

al-Dumur, Mahmud Awad Muslim

Thesis advisor

al-Abbadi, Muhammad Ali Husayn
al-Tarawinah, Raghad Muhammad

Comitee Members

al-Abadilah, Ahmad Hamad Hammud
al-Kasasibah, Muhammad Sharari Zamil
Sulayman, Hamzah Sabah Iyal

University

Mutah University

Faculty

Information Technology College

Department

Computer Science Department

University Country

Jordan

Degree

Master

Degree Date

2017

English Abstract

As the popularity of cloud computing has grown and with the continuous cheapening of computational resources there is a current trend of migration of services to cloud providers, capable of processing large amounts of data (large data) ,So the clustering in the cloud environments become a real process.

Clustering is a form of data modeling which is based on the construction of clusters.

Clusters are data sets that display the Property: the elements belonging to the same set that elements belonging to any other set, in a certain criterion of similarity.

Grouping of data in a cloud can be achieved using many different clustering techniques, this research attempts to evaluate the two clustering algorithms: Fuzzy K-mean data clustering algorithm (FKM) and K-mean data clustering algorithm (KM) in the cloud environment by adding them to the cloud computing simulator, to analyze the effect of some change factors on them, and finally comparing the results of the two added algorithms with each other and with the existing clustering algorithm in the simulator environment (vertical clustering algorithm VC & horizontal clustering algorithm HC).

The comparison factors used in the experiment are: the number of the data points to be clustered (large or small), the number of cluster and the type of dataset (random or ideal).

A cloud computing simulator that called work flowsim was used in the experiments that evaluate the algorithm’s behavior in different usage conditions.

FKM clustering algorithm gave close results to KM clustering algorithm, but it still requires more execution time than K-Means clustering, and the vertical clustering algorithm has the most execution time between all of the clustering algorithms (FKM, K-mean and HC) in all of the comparison cases.

Main Subjects

Information Technology and Computer Science

No. of Pages

45

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-780523