Metraipolis › archive › dc_68d585cb2083
Learning Sparse Decision Trees via Transformer Variational Auto-Encoders — arXiv:2609.01430
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.01430
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:34:19.702Z
- SHA-256 of the archived text
ab8f036f26ca5f4f3f3db7121dabd66a45bc4d51ff76be1e625e1cdb9f30b11f- Stable URL of this page
https://metraipolis.com/documento/dc_68d585cb2083
How to cite
Learning Sparse Decision Trees via Transformer Variational Auto-Encoders — arXiv:2609.01430 — arXiv (en), https://arxiv.org/abs/2609.01430 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_68d585cb2083 (Metraipolis archive dc_68d585cb2083, sha256 ab8f036f26ca5f4f3f3db7121dabd66a45bc4d51ff76be1e625e1cdb9f30b11f)
BibTeX
@misc{dc68d585cb2083,
title = {Learning Sparse Decision Trees via Transformer Variational Auto-Encoders — arXiv:2609.01430},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.01430},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_68d585cb2083; sha256 ab8f036f26ca5f4f3f3db7121dabd66a45bc4d51ff76be1e625e1cdb9f30b11f},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01430. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Decision trees are among the most widely used models in machine learning, largely due to their transparent decision logic, making them well-suited for high-stakes decision-making contexts. However, most existing learning algorithms focus on predictive performance, overlooking the joint optimization of other desirable properties, such as structural sparsity. In this work we propose TREVIS, an approach for learning decision trees with respect to complex objectives, based on the exploration of the latent space of a Tree Transformer Variational Auto-Encoder (TTVAE). By mapping decision trees onto latent representations, TREVIS replaces the discrete search space with a continuous one, enabling gradient-based optimization via a differentiable surrogate model. We experiment with TREVIS for learning decision trees that jointly optimize predictive performance and sparsity. Results show that TREVIS discovers decision trees matching the predictive performance of existing near-optimal algorithms while improving their structural sparsity. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.01430 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T06:15:12.898Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).