Mohammad A. Mezher
Email: mohd.mezher@brunel.ac.uk
PhD Supervisor: Dr. Maysam Abbod
Research Area: Intelligent Systems
Academic Qualification:
B.Sc Computer Science, Institute of Information and Communication Technology, Al-Zaytoonah University, Jordan
M.Sc Computer Science, Institute of Mathematics and Computer Science, Universiti Sains Malaysia (USM), Malaysia
Research Interests: Machine Learning, Statistical learning methods, Support Vector Machine, Evolutionary Algorithms, Biomedicine, Bioinformatics, Intelligent systems, Hybrid modelling systems.
PhD Research:
Support Vector Machines are a set of related supervised learning methods used widely for classification and regression problems. Kernel methods are learning systems which mapping nonlinear of the input data into linearly high dimensional feature spaces. The mapping feature is implemented via kernel functions which act as a dot product in reproducing kernel Hilbert spaces. Evolutionary algorithms have a promising result to draw a finest model selection for SVM. This generalization trick is implemented by substituting a kernel function for the dot products. This technique is useful whenever the kernel functions represent similarity between classification and regression data for more efficiency and high performances.
The main aims of this research is to develop new methods to solve problems in:
- Classification
- Regression
- Kernel function
- Model selection
Publication list
Journal Publications:
- Genetic Algorithm Self-Adaptive Mutation Rate for RNA Folding (GASAMR): (2009) Online Journal of Bioinformatics.
- Evolving Self-Adaptive Genetic Algorithm Using Nonlinear Support Vector Machines for Classification Problems: (2010) The International journal Annals Computer Science Series pp 99 - 112.
- Genetic Folding: A New Algorithm for Solving Multiclass SVM Problems: Applied Soft Computing journal, (Submitted).
- Genetic Folding: A New Class of Evolutionary Algorithms for SVM, IAJIT Journal (Submitted)
Conference Publications:
- Genetic Folding: A New Class of Evolutionary Algorithms for SVM: (2010) SGAI International Conference on Artificial Intelligence. Cambridge University.
- Genetic Folding Programming: A New Algorithm for Model Selection in SVM: JIEEEC 2011. (Submitted).
