Abstrait

Medical data mining algorithm based on improved rough set theory and probabilistic neural network

Zhang Qiu-ju, Li Jin-lin

As medical information system is popularized in more hospitals. Since it can collect more information about patientsÂ’ disease, it is feasible to use data mining technology to assist disease diagnosis. Based on rough set (RS) theory and PageRank algorithm, a new method was proposed to extract the key attributes of relevant attributes of diseases, and a probabilistic neural network (PNN) model was established for disease diagnosis. The results showed that the diagnostic accuracies of the model for patients with benign tumor and malignant tumor reached 100% and 95.24%, respectively, proving that the established model was effective and efficient in disease diagnosis.