Metraipolis › archive › dc_0b2c392146bb

Learning Conditional Expectation Operators via Functional Newton Updates — arXiv:2609.35598

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.35598
License (as archived)
CC0-1.0
Archived
2026-10-07T22:41:13.953Z
SHA-256 of the archived text
a595fe770c4aae1dfa5e9da385e73b6d0f3a5860cb2c23dd5150e54cdcaf0e39
Stable URL of this page
https://metraipolis.com/documento/dc_0b2c392146bb

How to cite

Learning Conditional Expectation Operators via Functional Newton Updates — arXiv:2609.35598 — arXiv (en), https://arxiv.org/abs/2609.35598 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0b2c392146bb (Metraipolis archive dc_0b2c392146bb, sha256 a595fe770c4aae1dfa5e9da385e73b6d0f3a5860cb2c23dd5150e54cdcaf0e39)
BibTeX
@misc{dc0b2c392146bb,
  title = {Learning Conditional Expectation Operators via Functional Newton Updates — arXiv:2609.35598},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35598},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0b2c392146bb; sha256 a595fe770c4aae1dfa5e9da385e73b6d0f3a5860cb2c23dd5150e54cdcaf0e39},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35598. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We introduce the Functional Spectral-Newton Method (FSNM) for learning the leading singular structure of a conditional expectation operator without fixing a basis or reproducing kernel Hilbert space. FSNM fits a low-rank representation of the centered joint-to-product density ratio kernel by alternating functional Newton updates. Each update reduces to a preconditioned regression, which we approximate with vector-valued regression trees in a stagewise boosting procedure. At the population level, we establish descent and an $O(1/T)$ best-iterate block-stationarity rate under a relative weak-learner accuracy condition, and show that every nondegenerate local minimum over the full centered $L^2$ spaces is a globally optimal rank-$d$ approximation. Synthetic experiments show that FSNM recovers a low-rank density ratio and its leading spectral structure, and that the same learned kernel can answer multiple conditional queries without refitting.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.LG).
Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.35598 — 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 2026-10-07T22:14:28.934Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).