Abstract:
Tertiary institutions in general have in recent years seen an upsurge in the number of students’ application and subsequent admittance into the various disciplines of the institutions.
Admission process has been a big concern for university administrations over the years where the influx of applications received on a yearly basis greatly tends to outnumber the available facilities in the university. This activity is a dynamic process that is carried out on a yearly basis, much more importantly ensuring that the prescribed quota as recommended by the Nigeria University Commission is not exceeded. Also, consideration has to be given to those from educationally less developed states, catchment area, supplementary and the likes. The admittance of competing students thus becomes a great challenge to the university administration.
This research examines how genetic algorithms can be used to optimize the network topology of neural networks. The research focuses on the application of the hybrid model of Artificial Neural Network and Genetic Algorithm a tool for optimization in the admission process of higher institutions of learning. Neural networks and genetic algorithms demonstrate powerful problem solving ability. They are based on quite simple principles, but take advantage of their mathematical nature: non-linear iteration.
Neural networks with back propagation learning provides results by searching for various kinds of functions. Genetic algorithm on the other hand is global search methods that are based on principles like selection, crossover and mutation. The technological approach for the development of the system is based on AMP (Apache, MySQL, and PHP) open source solution and was implemented using data obtained from The Federal University of Technology, Akure, Nigeria. The implementation results in optimal student admission processing system.
The result of the research showed genetic algorithm and neural network as an efficient
optimization tools which produced a very reliable results where classical data processing tool have been found inefficient.
