Intrusion detection model inspired by immune using K-means and naive Bayes as hybrid learning approach

مقدم أطروحة جامعية

Islim, Imad Faris Salim

مشرف أطروحة جامعية

Farhan, Hazim A.

أعضاء اللجنة

al-Tamimi, Abd al-Fattah
Naum, Riyad S.

الجامعة

جامعة الشرق الأوسط

الكلية

كلية تكنولوجيا المعلومات

القسم الأكاديمي

قسم علم الحاسوب

دولة الجامعة

الأردن

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2011

الملخص الإنجليزي

Intrusions in computer networks can be compared to human diseases with the difference that human body has an effective mechanism to deal with them.

Human immune system can detect and defend against unseen intruders.

Also, it is distributed and adaptive.

Human immune system is the most powerful defense system which may be helpful to apply its mechanism and properties in computer security field.

This thesis presents a model for intrusion detection system that consists of four components depending on innate/adaptive human immune system approaches and self/non-self theory of human immune system.

The proposed model is divided into two subsystems; the first one is attack response system which is similar to innate human immune system and the second is learning system which is similar to adaptive immune system.

Learning system is the core of the model; it presents a hybrid approach of machine learning through hybridization between k-Means clustering algorithm and Naive Bayes as a classifier.

The model goal is keeping information systems environment safe against intrusions and attacks through applying human immune system mechanism and properties to intrusion detection system.

Experimental results indecate that the proposed model provide a higher detection rate in both of DoS attacks and U2R attacks, which give the power to the proposed hybrid model and increase the security of information systems, especially in the critical environments.

التخصصات الرئيسية

تكنولوجيا المعلومات وعلم الحاسوب

عدد الصفحات

66

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-699437