Validating Symbolic Regression to Model Idiographic Symptom Dynamics

This proposal seeks to validate the Sparse Identification of Nonlinear Dynamics (SINDy) algorithm for modeling psychological time-series data. Study 1 tests SINDy’s ability to recover differential equations from formal theories of psychopathology, while Study 2 applies SINDy to real-world experience sampling data to uncover personalized computational models of symptom dynamics.