Application of Association Rule Algorithm in Chronic Disease Prediction

Chronic diseases, particularly chronic kidney diseases, have currently jeopardized people’s healthy life, with tens of thousands suffering from the disease each year. To improve the situation and, accordingly, to alleviate the caused harm, predictive research has attracted significant attention to itself. In line with this, among the numerous prediction research algorithms, Apriori is one of the most used ones. However, the algorithm suffers from a number of problems, including a reduction in its execution efficiency as well as the generation of more frequent item sets when the database is scanned for several times. As a consequence, more meaningless association rules are generated in predicting the disease. Therefore, for optimizing the Apriori association rules, the present study aims at proposing an algorithm based on pointer array and difference set theory. The results showed that the use of pointer array led to the quick reduction of memory occupation and the used time for intersection, which in turn, improved the execution efficiency of the algorithm. Moreover, experiments were carried out to verify different support thresholds and transaction records, the obtained results of which demonstrated a great improvement in memory space occupancy and execution time. Finally, the algorithm was applied to predict the chronic diseases, the results of which realized the probability prediction to a certain extent, provided a reference for specialists to diagnose the diseases, and was significant for the promotion of automated medical treatment.