UTHM Institutional Repository

Applying GA as a method of generating class schedule. A case study in UTHM

Wan Ismail, Wan Zilliani (2013) Applying GA as a method of generating class schedule. A case study in UTHM. Masters thesis, Universiti Tun Hussein Onn Malaysia.

[img]
Preview
PDF
WAN_ZILLIANI_BINTI_WAN_ISMAIL.pdf

Download (1MB)

Abstract

University Class Timetabling Problem (UCTP) is a common problem faced any regular educational institutions. This problem encountered as it has multivariable parameter combinations and multiple constrains. This project focuses on the concept of Genetic Algorithm which is an alternative solution to the class of multi-constrained, NP-hard, combinatorial optimization problems. The parameter involved is time slots, classes, courses, lecturers, distances and constraints among it all. Genetic Algorithm revolves around process operator like evaluation, crossover, selection and mutation. All these process are designed to generate multiple solutions which near to the best solution. However, the perfect solution did not exist. Past work have been done on the same matter with different approaches such as Mathematical Model, Artificial Bee Colony Algorithm, Memetic Algorithm, Multiswap Algorithm, Ant Colony System, Immune Network System and Hybrid Genetic Algorithm. However, this paper uses Simple Genetic Algorithm approach which manipulates the GA operator to achieve the objectives. The data been divided into two part which is variables and fixed data and encoded to binary data to ease the processing. All the data designed in array called chromosomes and randomized to create initial populations. The population been evaluated to select the best solutions by filtering the high points achieved by each chromosome. The successful individuals will go through crossover, selection and mutation process to generate new population with the possible best solution. However, the lecturer constrain faced some difficulties to fit in the program and being removed. The program can only accommodate 17 modules for 300 initial populations generated within 10 rounds which ordered a 199 successful individual with score 500 above. Extra modules can be processed if the number of initial population reduced.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > QA76 Computer software
Depositing User: Normajihan Abd. Rahman
Date Deposited: 24 Jun 2014 04:46
Last Modified: 24 Jun 2014 04:46
URI: http://eprints.uthm.edu.my/id/eprint/5552
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year