Classification Performance Enhancement for Students Realisation Model

Classification Performance Enhancement for Students Realisation Model
                                             

Tarik A. Rashid


Software Engineering, College of Engineering, Hawler, Kurdistan, Iraq

E_mail : tarik.rashid@su.edu.krd





Article info


Original: 20 Apr. 2015
Revised: 6 June 2015
Accepted: 25 June 2015
Published online:  20 Sep. 2015  



Key Words:

Forecasting Student Performance, 
Feature Reduction, 
Neural Networks, 
Support Vector Machines, 
K-Nearest Neighbors, 
Genetic Algorithms.






Abstract
This research work aims at enhancing a classification task for student’s realisation model at Salahadin University, Hawler, Kurdistan.  1000 records of data from different colleges and departments at Salahadin University are collected to conduct this research work. The collected data has been pre-processed, cleaned, filtered, normalized, then after, feature selection techniques are applied to reduce the dimensionally of the data,  finally a classification task is carried out to find the realization of students. The results show that a model of Support Vector Machine +Genetic Algorithm + Artificial Neural Network produces promising results than other models.





Ċ
Kewan Omer,
Sep 20, 2015, 12:07 PM