A Hierarchical Model for Association Rule Mining of Sequential Events: An Approach to Automated Medical Symptom Prediction

摘要: In many healthcare settings, patients visit professionals periodically and report multiple medical conditions, or symptoms, at each encounter. We propose a statistical modeling technique, called the Hierarchical Association Rule Model (HARM), that predicts patient's possible future symptoms given current past history of reported symptoms. The core our technique is Bayesian hierarchical model for selecting predictive association rules (such as "symptom 1 symptom 2 → 3") from large set candidate rules. Because this method "borrows strength" using similar patients, it able to provide predictions specialized any patient, even when little information about available.