Machine learning and data-driven models

Machine learning and data-driven models rely on patterns in the data to answer research questions, as compared to relying on researcher-generated hypotheses. These data-driven models show great promise in helping the field predict very difficult-to-predict outcomes, like eating disorder severity and suicide.

Example questions I tackle within this area include:

1. How can we best define eating disorder severity?
2. Does a better definition of eating disorder severity relate to or improve treatment outcomes?
3. How can we better predict eating disorder treatment outcomes?

Example papers within this content area include, but are not limited to:

Forrest, L. N., Ivezaj, V., & Grilo, C. M. (2023). Machine learning versus traditional regression models predicting treatment outcomes for binge-eating disorder from a randomized controlled trial. Psychological Medicine, 53, 2777–2788.

Forrest, L. N., Jacobucci, R. C., & Grilo, C. M. (2022). Empirically determined severity levels for binge-eating disorder outperform existing severity classification schemes. Psychological Medicine, 52, 685–695.

Ortiz, S. N., Forrest, L. N., Ram, S. R., Jacobucci, R. C., & Smith, A. R. (2021). Using shape and weight overvaluation to empirically differentiate severity of other specified feeding or eating disorder. Journal of Affective Disorders, 295, 446–452.