Machine Learning Model Identifies Psychological Resilience Levels in Health Care Workers With 75 Percent Accuracy

September 22, 2026 - 00:20

A study published in Discover Artificial Intelligence used logistic regression, random forest and support vector machine (SVM) to classify health care workers according to their level of psychosocial resilience. The researchers reported that the best performing model reached roughly 75 percent accuracy, suggesting that machine learning can serve as a practical tool for spotting employees who may need support before stress and burnout take a deeper toll.

The work focused on resilience as a measurable psychological trait rather than a vague idea. Health care staff face long shifts, heavy patient loads and emotionally draining situations, all of which raise the risk of anxiety, depression and exhaustion. Being able to sort workers into resilience groups could help hospitals direct limited mental health resources toward those who need them most.

Three algorithms were compared during the analysis. Logistic regression offered a simple and interpretable baseline, random forest captured complex interactions among variables, and the SVM searched for boundaries between groups in the data. The SVM produced the strongest results, though the study noted that each method has tradeoffs in transparency and ease of use.

The authors stressed that 75 percent accuracy is useful but not perfect. A model of this kind should support human judgment, not replace it. They also pointed to the need for larger samples and validation across different hospitals and regions before such tools become routine. Even so, the findings add to growing evidence that data driven methods can help protect the well being of the people who care for others.