2nd Place Winner of Fall 2022
by Team ML Queens: Katya Mikhailova, Wamia Said, Katherine Cadzow
Abstract: As COVID-19 mutates and develops resistance against outdated vaccinations, it is more important than ever to stay up to date with COVID vaccine booster shots. This study seeks to take data from the Census Bureau’s House Pulse Survey to better understand which populations are least likely to get their boosters. These results will, in turn, aid us in effectively allocating resources towards groups that are in greatest need of being boosted. To do this, we applied several machine learning algorithms that enabled us to predict whether an individual is boosted based on their responses to the rest of the Household Pulse Survey. We used a Decision Tree, Random Forest, and SVM classifiers, and after analyzing our results, we came to the conclusion that the Decision Tree model produced the most satisfactory results in the context of our problem. The Decision Tree model provided us with valuable insight into the six most important features for predicting booster status from our dataset: birth year, expense difficulty, education level, number of non-booster COVID vaccine doses, price-related stress, and COVID status