Machine Learning Applications in Psychotherapy Research
Persons involved: Anna Georg, Oliver Evers, Max Zettl, Paul Schröder-Pfeifer.
Machine learning approaches for psychotherapy research have the potential to use statistical algorithms to examine multidimensional data sets for non-linear relationships and interactions and thereby generate generalizable predictions at the individual level. However, this type of data analysis is still relatively new in psychotherapy research and thus there is still little knowledge about how different algorithms and validation procedures behave in data sets of different sizes and structures. To address this gap, we at the Institute for Psychosocial Prevention are using machine learning in a variety of projects, from diagnosing personality disorders to searching for risk variables for parental stress.
Duration: 2017 – ongoing