Adaptive Apriori and Weighted Association Rule Mining on Visual Inspected Variables for Predicting Obstructive Sleep Apnea
Doreen Ying Ying Sim, Chee Siong Teh, Ahmad Izuanuddin Ismail · 2014
A prediction model for Obstructive Sleep Apnea (OSA) is developed based on a novel formulation approach by using customized Associative Rule (AR) Mining Techniques, i.e. Adaptive Apriori (AA) and Weighted Association Rule Mining (WARM), on visual inspected variables. This prediction model is based on the typical clinical data sets obtained from several hospitals where from derivation of association rule mining (data-driven) techniques and medical knowledge on these variables (knowledge-driven) were applied separately to our training data sets. Application of our prediction framework to our testing data sets showed a significant improvement in terms of prediction accuracy and level of efficiency as compared with the classical approach of using medical Experts’ Rules (ERs).