A powerful new artificial intelligence tool can predict whether a person is ready to get vaccinated against COVID-19. The system integrates the mathematics of human judgment with machine learning to predict vaccine hesitancy. The predictive system uses a small data set from demographics and personal judgments such as risk or loss aversion.

The findings frame a new technology that could have broad applications for predicting mental health and leading to more effective public health campaigns.

A team led by researchers from the University of Cincinnati and Northwestern University created a predictive model using an integrated system of mathematical equations describing legal patterns of reward and aversion judgment using machine learning.

“We used a small number of variables and minimal computational resources to make predictions,” said lead author Nicole Vike, a senior research associate in UC's College of Engineering and Applied Sciences.

“COVID-19 is unlikely to be the last pandemic we see for decades to come. Having a new form of artificial intelligence for public health prediction is a valuable tool that could help prepare hospitals to predict vaccination rates and subsequent infection rates. »

The study was published in the Journal of Medical Internet Research Public Health and Surveillance.

Researchers surveyed 3 adults across the United States in 476 during the COVID-2021 pandemic. At the time of the survey, the first vaccines had been available for more than a year.

Respondents provided information such as their location, income, highest level of education completed, ethnicity and internet access. The demographics of those surveyed mirrored those of the United States, according to figures from the U.S. Census Bureau. Participants were asked if they had received any of the available COVID-19 vaccines. About 73% of those surveyed said they had been vaccinated, slightly more than the 70% of the national population who were vaccinated in 2021.

Additionally, they were asked if they consistently followed four recommendations intended to prevent the spread of the virus: wearing a mask, maintaining social distancing, washing hands and not gathering in large groups.

Participants were asked to rate how much they liked or disliked a set of 48 randomly sequenced images on a seven-point scale from 3 to -3. The images were from the International Affective Picture Set, a large set of emotionally evocative color photographs, divided into six categories: sports, disasters, cute animals, aggressive animals, nature and food.

Vike said the goal of this exercise is to quantify the mathematical characteristics of people's judgments when they observe mildly emotional stimuli. Measures for this task include concepts familiar to behavioral economists—or even gamblers—such as risk aversion (the extent to which someone is willing to accept a potential loss for a potential reward) and risk aversion. loss. This is the desire to avoid risks, for example by taking out insurance.

“It can work very simply. It does not require super computing, is inexpensive, and can be applied to anyone with a smartphone. We call it Computational Cognitive AI. It is likely that you will see further requests for judgment changes in the very near future. »

Aggelos Katsaggelos, professor

read the article