By

First and foremost, I am greatly indebted to the supervisor of this work, Prof. Amir Atiya, for his continuous support. We would like to acknowledge Prof. Nevin Darwish, and Prof. Magda Fayek (Cairo University); and Dr Pierre-Emmanuel Hladik, and Prof. Patrick Senac (LAAS CNRS) for the useful comments they provided. We specially thank Eng. Wael Nabil for preparing smart animations that have helped a lot in illustrating our proposed idea. Last, but not least, I would like to thank my mother for her patience and endless encouragement. i In this thesis, we propose the so-called ”SVM’ed-kernel function ” and its use in SVM classification problems. This kernel function is itself a support vector machine classi-fier that is learned statistically from data. We show that the new kernel manages to change the classical methodology of defining a feature vector for each pattern. One will only need to define features representing the similarity between two patterns al-lowing many details to be captured in a concise way. The new proposed kernel show