Year: 2026
Type: article
Abstract: We introduce ordinal mixed-effect random forests (OMERF), a novel statistical method that extends random forests to handle hierarchical data with ordinal responses. OMERF retains the flexibility of tree-based ensemble methods for modeling complex relationships among categorical and continuous variables, while explicitly accounting for hierarchical dependence structures, enabling statistical inference across all data levels. A simulation study is conducted to validate the performance of the proposed method and to compare it to those of other state-of-the-art models. The application of OMERF is exemplified in two case studies. The former focuses on predicting students’ performance using data from the Programme for International Student Assessment (PISA) 2022, while the latter analyzes individual life satisfaction across European countries using data from the European Social Survey (ESS).
Source: Journal of Classification
Authors: Giulia Bergonzoli, Lidia Rossi, Chiara Masci
Institutions: European Central Bank, Politecnico di Milano, University of Milan
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,290
Funder: Università degli Studi di Milano
OpenAlex ID: W7212173663