Implementation and Comparison of Inductive Learning Algorithms on Timetabling

Academic timetabling has been known as a complex task due to various rules that needs to be satisfied. In order to build an effective and efficient timetabling system, these rules have to be identified and recognized. This paper focuses on the implementation and comparison of inductive learning algorithms on timetabling for the purpose to see the ability of several inductive learning algorithms towards solving complex problems, the ability of the algorithms to adapt to the new environment by changing rules and to compare the performance between the algorithms. The algorithms used for this study are ID3 algorithm, AQ algorithm and ILA algorithm. The study shows that all three inductive learning algorithms were successful in the implementation and able to adapt to new environment. ILA algorithm is shown to have better performance in the number of rules generated and percentage of accuracy compared to ID3 algorithm and AQ algorithm. Keyword: Inductive Learning, Academic Timetabling, ID3 algorithm, AQ algorithm and ILA algorithm.