Global Interpretability via Automated Preprocessing: A Framework Inspired by Psychiatric Questionnaires

arXiv:2602.23459v2 Announce Type: replace Abstract: Psychiatric questionnaires are highly context sensitive and often only weakly predict subsequent symptom severity, which makes the prognostic relationship difficult to learn. Although flexible nonlinear models can improve predictive accuracy, their limited interpretability can erode clinical trust. In fields such as imaging and omics, investigators commonly address visit- and instrument-specific artifacts by extracting stable signal through preprocessing and then fitting an interpretable linear model. We adopt the same strategy for questionnaire data by decoupling preprocessing from prediction: REFINE (Redundancy-Exploiting Follow-up-Informed Nonlinear Enhancement) uses follow-up information to learn a nonlinear, item-aligned preprocessing of baseline measurements, and then maps these stabilized baseline items to future severity through a linear coefficient matrix. This construction constrains where nonlinearity enters rather than restricting the end-to-end predictive function class, allowing the prognostic relationship to remain globally interpretable through a single coefficient matrix rather than through post hoc local attributions. In the population, the resulting predictor recovers the Bayes-optimal conditional mean, so global linear interpretability does not require sacrificing nonlinear predictive flexibility. In experiments, REFINE outperforms other interpretable approaches while preserving clear global attribution of prognostic factors across psychiatric and non-psychiatric longitudinal prediction tasks.

Related

• Deep learning with missing data
• Did Models Learn Sufficiently? Attribution-Guided Training via Subset-Selected Counterfactual Augmentation
• Majorization-Minimization Networks for Inverse Problems: An Application to EEG Imaging
• SpecTM: Spectral Targeted Masking for Trustworthy Foundation Models

Source: arXiv cs.LG | 2026-08-17