Abstract:
Understanding the factors to identify the optimal COVID-19 vaccine is essential
for maximizing its impact and is crucial to vaccination success. Some
adverse side effects were observed during vaccination due to vaccine selection
without analyzing the patient’s medical history. The correlation between patient
medical history and the side effects of the vaccine should be addressed.
This project proposes a novel approach to predict the optimal COVID-19 vaccine
for a person based on their medical history and analyzing the vaccine side
effects caused by each disease found in their medical history. Three different
machine learning models (SVM, CNN, and KNN) were trained with medical
history and side effects after vaccination of 4153 people from various parts of
Bangladesh. This system proposes a weighted scoring system for each COVID-
19 vaccine based on its side effects after vaccination, correlating the patient’s
medical history. The proposed system achieved accuracies of 54.01%, 48.49%,
and 53.77% for SVM, CNN, and KNN, respectively. The proposed system will
play an important role in selecting the most effective and suitable COVID-19
vaccine for people with severe health risks.
