Splitting Stump Forests: Tree Ensemble Compression for Edge Devices (extended Version)

Abstract

Abstract We introduce Splitting Stump Forests—small ensembles of weak learners extracted from a trained random forest. The high memory consumption of random forests renders them unfit for resource-constrained devices. We show empirically that we can significantly reduce the model size and inference time by selecting nodes that evenly split the arriving training data and applying a linear model on the resulting representation. Our extensive empirical evaluation indicates that Splitting Stump Forests outperform random forests and state-of-the-art compression methods on memory-limited embedded devices.

Cite

Text

Alkhoury et al. "Splitting Stump Forests: Tree Ensemble Compression for Edge Devices (extended Version)." Machine Learning, 2025. doi:10.1007/S10994-025-06866-2

Markdown

[Alkhoury et al. "Splitting Stump Forests: Tree Ensemble Compression for Edge Devices (extended Version)." Machine Learning, 2025.](https://mlanthology.org/mlj/2025/alkhoury2025mlj-splitting/) doi:10.1007/S10994-025-06866-2

BibTeX

@article{alkhoury2025mlj-splitting,
  title     = {{Splitting Stump Forests: Tree Ensemble Compression for Edge Devices (extended Version)}},
  author    = {Alkhoury, Fouad and Buschjäger, Sebastian and Welke, Pascal},
  journal   = {Machine Learning},
  year      = {2025},
  pages     = {219},
  doi       = {10.1007/S10994-025-06866-2},
  volume    = {114},
  url       = {https://mlanthology.org/mlj/2025/alkhoury2025mlj-splitting/}
}