A Novel Recommendation System for Vaccines Using Hybrid Machine Learning Model

Advancements in the medical field have brought almost all diseases within the realm of vaccinology. Multiple manufacturers produce vaccines for various diseases. The efficacy rate of these vaccines varies based on different factors. Over the years researchers have developed various recommendation systems for drugs using different techniques with the common goal to help doctors in prescribing drugs after considering different factors. The recommendation system for vaccines is an unexplored area that requires extensive consideration of factors to recommend the vaccine that provides a high efficacy rate. In our paper, we propose a recommendation system for vaccines that considers host-based factors such as age, sex, medical history and also, vaccine-based factors such as post-vaccination recovery rate, death rate and symptoms faced by recipients. The algorithm is score-based where each vaccine is given a score and ranked accordingly for the patient. We created a hybrid machine learning model to extract useful information from the medical data to score the vaccines that are suitable for a particular recipient. The results produced by the system when run on simulated patient data show significant changes in recommended vaccines for the recipient based on different factors.