Researchers applied three machine learning approaches to data gathered from healthcare workers during the COVID-19 pandemic, aiming to classify individuals by their level of psychosocial resilience. The study, published in Discover Artificial Intelligence, reports that the models achieved 75% accuracy in distinguishing between resilience levels.
The algorithms used were logistic regression, random forest, and support vector machine. Each was trained on information collected during the pandemic, though the specific features used for classification are not detailed in the available summary.
The finding suggests that machine learning could help identify healthcare workers who may need additional psychological support. However, the 75% accuracy rate also means a quarter of classifications were incorrect, so further validation would be needed before the approach could be used in real-world settings.