A Epilepsy Drug Recommendation System by Implicit Feedback and Crossing Recommendation

Epilepsy refers to a set of chronic neurological syndromes characterized by transient and unexpected electrical disturbances of the brain. It is estimated that there are about 65 million patients in globe suffering from epilepsy and epilepsy, if not impossible, is extremely hard to cure eventually. However, approximately 70% of seizures can be under control by drugs. Electronic Health Records (EHRs) of epileptics are very important data resources for personalized medicine prescription. In this paper, we take real medical electronic cases to conduct large data analysis, and propose a drug recommendation system by Implicit Feedback and Crossing Recommendation (IFCR) to help doctors choose drugs. The proposed system aims to investigate epileptics' medical history in order to find the relationships between the syndromes and the drugs. Compared with a baseline system using Artificial Neural Network (ANN), our proposed system performs much better than ANN in terms of the recall rate with up to 30% improvement. In general, the performance of IFCR is better than that of ANN. Finally, we analyze the recommendation results of two algorithms and discover it is possible to propose an ensemble model to combine IFCR with ANN to exploit their respective advantages in drug recommendation.