Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2605.26653 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916057728417792 |
|---|---|
| author | Manage, Sithija Wang, Y. Samuel Wells, Martin T. |
| author_facet | Manage, Sithija Wang, Y. Samuel Wells, Martin T. |
| contents | In regression problems where covariates are naturally organized in a hierarchical tree structure, a central challenge is to select the resolution at which covariates enter the model. Determining this level of feature aggregation is of intrinsic scientific interest and can improve statistical efficiency by inducing sparsity. While a rich literature addresses this problem in the linear setting, extending feature aggregation to the nonlinear setting remains an open challenge. In this work, we propose to simultaneously perform model selection and feature aggregation through a penalized Nadaraya-Watson-type estimator. Our proposed estimator, Kernel Regression with Tree-EXploring AggregationS (KR-TEXAS), constructs adaptive penalty weights for the features based on pilot estimators of the regression function's partial derivatives. Under mild conditions, we establish model selection consistency for a well-defined target aggregation set, and our simulations show strong performance in both model selection and prediction. Finally, we demonstrate the utility of our procedure by applying it to a microbiome data set to predict short chain fatty acids. A user-friendly implementation of our procedure is available in the R package krtexas. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_26653 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Nonparametric Regression via Tree-Guided Feature Aggregation Manage, Sithija Wang, Y. Samuel Wells, Martin T. Methodology In regression problems where covariates are naturally organized in a hierarchical tree structure, a central challenge is to select the resolution at which covariates enter the model. Determining this level of feature aggregation is of intrinsic scientific interest and can improve statistical efficiency by inducing sparsity. While a rich literature addresses this problem in the linear setting, extending feature aggregation to the nonlinear setting remains an open challenge. In this work, we propose to simultaneously perform model selection and feature aggregation through a penalized Nadaraya-Watson-type estimator. Our proposed estimator, Kernel Regression with Tree-EXploring AggregationS (KR-TEXAS), constructs adaptive penalty weights for the features based on pilot estimators of the regression function's partial derivatives. Under mild conditions, we establish model selection consistency for a well-defined target aggregation set, and our simulations show strong performance in both model selection and prediction. Finally, we demonstrate the utility of our procedure by applying it to a microbiome data set to predict short chain fatty acids. A user-friendly implementation of our procedure is available in the R package krtexas. |
| title | Nonparametric Regression via Tree-Guided Feature Aggregation |
| topic | Methodology |
| url | https://arxiv.org/abs/2605.26653 |