Basic Informations

C.V

• Full Name, Mohammed Abo Bakr Ibrahim Mohammed El-Ashiri
• Nationality, Egyptian
• Marital Status, Married (3 Ch.)
• D.Birth, 18, January 1977.
• Languages, Arabic and English
• Job title (Egypt), Lecturer [July. 2012 - present], Computer Sciences Department, Lecturer Assistant  in Computer, Faculty of Computer and Information system , Beni Suef University, Egypt 

Master Title

Fuzzy Decision Trees Approach for Data Mining

Master Abstract

Decision maker is quite interested to find out that certain patterns that exist and control of the data stored in the database. Data mining is the powerful scientific approach that can be used to find out hidden patterns. One of the most famous techniques of data mining is decision trees, which used for generating trees and extracting a set of rules. Traditional decision trees for a huge database may not be helpful and inability to generate decision tree from group of imprecise training examples. The objective of the thesis is to combine decision trees and fuzzy set theory to develop new technique of data mining to overcome the disadvantages of crisp decision trees and fuzzy decision trees, such as inability to generate trees with smaller size and higher accuracy directly than without pruning algorithms This thesis presents a new algorithm for combining both of “Ambiguity measure” and “Classifiability measure.

PHD Title

A Novel Approach for Integrating Decision Trees, Fuzzy Sets and Rough Set

PHD Abstract

A decision maker is not usually interested in making queries in several thousands of records. However, a decision maker is quite interested to find out that certain hidden patterns that exist and control most of the data stored in the database. Data Mining is the powerful scientific approach that can be used to find out such hidden patterns. One of the most famous techniques of data mining is decision trees. Decision trees are widely used for generating trees and extracting a set of -IF THEN- rules, which are quite helpful for making decisions. Classical searching techniques "Traditional Decision Trees" for a huge database may not be helpful for effective decision-making processes. The main serious problem of decision trees is that very large trees may be generated and would added nothing to the process of the decision-making. The main objective of the thesis is to combine the decision trees and fuzzy set theory with rough set theory to develop a new technique for Data Mining to overcome the disadvantages of traditional decision trees as well as existing fuzzy decision trees. Several algorithms of fuzzy decision trees such as FID3, fuzzy Ambiguity and fuzzy decision trees using Dependency Degree are studied to develop our proposed algorithm. The proposed algorithm is found to be at least equal or higher in accuracy and smaller in size when compared with other algorithms.