Enhancing associative classification by combining simulated annealing with genetic algorithm

مقدم أطروحة جامعية

Najib, Muadh Mustafa Ahmad

مشرف أطروحة جامعية

al-Shaykh, Asim A. R.

أعضاء اللجنة

Abu al-Suud, Salih Mustafa
Kanan, Raid Karim
Hattab, Izz al-Din Shakir Hasan

الجامعة

الأكاديمية العربية للعلوم المالية و المصرفية

الكلية

كلية نظم و تكنولوجيا المعلومات

القسم الأكاديمي

قسم نظم المعلومات الحاسوبية

دولة الجامعة

الأردن

الدرجة العلمية

دكتوراه

تاريخ الدرجة العلمية

2011

الملخص الإنجليزي

Association Rule Mining (ARM) is a technique in Data Mining (DM) that aims to discover frequent rule items from database, the integration between association rule mining and classification initiates an alternative classification approach called Associative Classification (AC).

A lot of AC algorithms proposed in literature intended to enhance the accuracy of the classification.

In this thesis we propose three novel models to enhance the AC accuracy; we use two Artificial Intelligence (AI) techniques which are Simulated Annealing (SA) and Genetic Algorithm (GA) as optimization methods to serve our purpose.

The first and second models adds a new phase called "Classifier Arrangement" to Associative Classification structure by employing different Artificial Intelligence techniques, the aim if these models is to get the best order of the final rules in the classifier.

The third model proposes a new Rule Ranking method by employing a hybrid Artificial Intelligence (AI) technique that combines Simulated Annealing (SA) with Genetic Algorithm (GA).

The experiments of our models are conducted on various benchmark data sets from UCI Machine Learning Repository and the results show that our models enhance clearly the AC accuracy.

التخصصات الرئيسية

تكنولوجيا المعلومات وعلم الحاسوب

الموضوعات

• الخوارزميات
• معالجة البيانات
• التنقيب في البيانات

عدد الصفحات

120

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-306729