Order Dependence in Regression by Composition: Discussion on "Regression by Composition'' by Farewell, Daniel, Stensrud, and Huitfeldt

We discuss the regression-by-composition framework of Farewell, Daniel, Stensrud and Huitfeldt, highlighting a key consequence of its sequential construction: order dependence. Reordering the flows may change the implied conditional distribution, the interpretation of model parameters, and the associated estimation problem, with consequences for model specification, interpretation, and inference.

Bayesian Methods and Mixture Models Statistical Distribution Estimation and Applications Statistical Methods and Bayesian Inference

Paper

Similar papers

Sequential Dependence Decomposition and Feature LearningXiangxiang Xu, Lizhong Zheng2023 In Library

Compositional Linear Regression on Interval-valued DataDirenc Pekaslan, Christian Wagner2021 · IEEE Symposium Series on Computational Intelligence · 2 citations In Library

Distribution-Free Inference Beyond ExchangeabilitySÉRGIO DE ANDRADE, PAULO2025 · Zenodo (CERN European Organization for Nuclear Research) In Library

Distribution-Free Inference Beyond ExchangeabilitySÉRGIO DE ANDRADE, PAULO2025 · Zenodo (CERN European Organization for Nuclear Research) In Library

Distribution-Free Inference Beyond ExchangeabilitySÉRGIO DE ANDRADE, PAULO2025 · Zenodo (CERN European Organization for Nuclear Research) In Library

Foundations of compositional models: inferenceVl. Bína, Radim Jiroušek, Václav Kratochvíl2021 · International Journal of General Systems In Library

Compositional Generalization in Autoregressive Models via Logit CompositionAakash Kumar, Maria Sofia Bucarelli, Emanuele Natale2026 · arXiv.org · 0 citations In Library

Compositional Model Repositories via Dynamic Constraint Satisfaction with Order-of-Magnitude PreferencesJeroen Keppens, Q. Shen2011 · Journal of Artificial Intelligence Research · 17 citations In Library

Technical Report: Partial Dependence through StratificationT. Parr, James D. Wilson2019 · 3 citations In Library

Copy-composition for Probabilistic Graphical ModelsToby St Clere Smithe2024 · arXiv (Cornell University) · 5 citations In Library

Source of record

OpenAlex

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal