Artificial representative trees provide interpretable stable predictions with uncertainty
Beyond Point Prediction: Artificial Representative Trees with Uncertainty
Machine Learning
Summary
Random forests are good at making predictions but hard to understand, while single decision trees are easy to understand but less reliable. The authors worked on a kind of decision tree called artificial representative trees (ARTs) that can mimic random forests and also estimate uncertainty in predictions. They combined ARTs with a method called conformal predictive systems to make predictions with confidence intervals and probabilities. Their approach created stable and easy-to-understand trees that performed nearly as well as standard decision trees when tested on many datasets.
What this means in practice
- •For hospital data teams: Build interpretable predictive models with reliable uncertainty estimates to support medical decision-making using patient data.
- •For financial risk analysts: Create stable and transparent prediction models that provide calibrated probabilities and intervals for risk assessment tasks.
Authors
Lea L. Mairhöfer, Silke Szymczak, Björn-Hergen Laabs, Tuwe Löfström-Cavallin
Abstract
Random forests (RFs) predict well but are opaque, whereas single decision trees are interpretable but unstable. Artificial representative trees (ARTs) were developed as interpretable surrogate models for RFs, but their use as standalone prediction models with uncertainty quantification has not been systematically investigated. We combine ARTs with leaf-wise Mondrian conformal predictive systems (CPS), enabling a single tree to provide continuous predictions, prediction intervals, and probabilities of exceeding arbitrary thresholds. We compared ARTs with CPS against decision trees with CPS and separate regression and probability trees across five simulation scenarios, 21 benchmark datasets, and a cross-sectional NHANES example data set. Repeated cross-validation assessed predictive performance, interpretability, and stability. ARTs with CPS yield compact, structurally stable trees with substantially more reproducible split-variable selection than decision trees across benchmark datasets and NHANES. Decision trees showed slightly better predictive performance and narrower prediction intervals, while coverage was broadly comparable. CPS-based trees generally achieved lower and less variable Brier scores than multi-model approaches. Combining ARTs with CPS therefore provides a single, interpretable, and stable model for continuous predictions and calibrated probabilities, balancing predictive performance with reproducibility and transparency in settings where stability and interpretability are essential.