Finding Patterns in Medical Ward Data Using Rough Sets

摘要

Data have been obtained from a hospital in Saudi Arabia. In this project, we are discovering patterns an experimental tool called Rough Set Graphic User Interface (RSGUI). Several algorithms are available in RSGUI, each of which is based in Rough Set theory. Our objective is to find short meaningful predictive rules. First we find a minimum set of attributes that fully characterize the data. Some of the rules generated from this minimum set of attributes were obvious and therefore uninteresting. Others were surprising and therefore interesting. RSGUI allows fine-tuning of rules making it possible to analyze information about the rules and compare the rules resulting from different algorithms. Usual measures of strength of the rule, such as length of the rule, certainty and coverage were considered. In addition, a measure of interestingness of the rules has been developed based on questionnaires administered to human subjects.