Healthcare Providers Recommender System Based on Collaborative Filtering Techniques
Abdelaaziz Hessane, Ahmed El Youssefi, Yousef Farhaoui, Badraddine Aghoutane, Noureddine Ait Ali, Ayasha Malik · 2022
Finding a suitable health care provider (physicians, clinics, therapeutic centers, etc.) seems to be a time-consuming and difficult task. Before deciding, the patient should gather input from others who are in a similar situation. The latter can be determined in a variety of ways, such as by looking for individuals who have similar symptoms or who have similar attitudes toward a group of health professionals. Both ways can lead to overchoice problem, consequently, the patient can't make the right decision. In this chapter, we try to automate this process of finding a healthcare provider to solve the above-mentioned problems through the technique of recommendation systems, one of the applications of machine learning (ML). To create a predictive patient-healthcare provider rating model, multiple collaborative filtering strategies were introduced, tested, and evaluated. The algorithms for sparse matrix completions were implemented in four different ways: k-nearest neighbors (KNN) based, matrix factorization, slop one, and co-clustering. The root mean squared error (RMSE) metric was used to compare and evaluate the accuracy of the different models.