Simplex-FEM Networks (SiFEN): Learning A Triangulated Function Approximator

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yahyati, Chaymae, Lamaakal, Ismail, Makkaoui, Khalid El, Ouahbi, Ibrahim, Maleh, Yassine
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917190422233088
author Yahyati, Chaymae
Lamaakal, Ismail
Makkaoui, Khalid El
Ouahbi, Ibrahim
Maleh, Yassine
author_facet Yahyati, Chaymae
Lamaakal, Ismail
Makkaoui, Khalid El
Ouahbi, Ibrahim
Maleh, Yassine
contents We introduce Simplex-FEM Networks (SiFEN), a learned piecewise-polynomial predictor that represents f: R^d -> R^k as a globally C^r finite-element field on a learned simplicial mesh in an optionally warped input space. Each query activates exactly one simplex and at most d+1 basis functions via barycentric coordinates, yielding explicit locality, controllable smoothness, and cache-friendly sparsity. SiFEN pairs degree-m Bernstein-Bezier polynomials with a light invertible warp and trains end-to-end with shape regularization, semi-discrete OT coverage, and differentiable edge flips. Under standard shape-regularity and bi-Lipschitz warp assumptions, SiFEN achieves the classic FEM approximation rate M^(-m/d) with M mesh vertices. Empirically, on synthetic approximation tasks, tabular regression/classification, and as a drop-in head on compact CNNs, SiFEN matches or surpasses MLPs and KANs at matched parameter budgets, improves calibration (lower ECE/Brier), and reduces inference latency due to geometric locality. These properties make SiFEN a compact, interpretable, and theoretically grounded alternative to dense MLPs and edge-spline networks.
format Preprint
id arxiv_https___arxiv_org_abs_2511_04804
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Simplex-FEM Networks (SiFEN): Learning A Triangulated Function Approximator
Yahyati, Chaymae
Lamaakal, Ismail
Makkaoui, Khalid El
Ouahbi, Ibrahim
Maleh, Yassine
Machine Learning
We introduce Simplex-FEM Networks (SiFEN), a learned piecewise-polynomial predictor that represents f: R^d -> R^k as a globally C^r finite-element field on a learned simplicial mesh in an optionally warped input space. Each query activates exactly one simplex and at most d+1 basis functions via barycentric coordinates, yielding explicit locality, controllable smoothness, and cache-friendly sparsity. SiFEN pairs degree-m Bernstein-Bezier polynomials with a light invertible warp and trains end-to-end with shape regularization, semi-discrete OT coverage, and differentiable edge flips. Under standard shape-regularity and bi-Lipschitz warp assumptions, SiFEN achieves the classic FEM approximation rate M^(-m/d) with M mesh vertices. Empirically, on synthetic approximation tasks, tabular regression/classification, and as a drop-in head on compact CNNs, SiFEN matches or surpasses MLPs and KANs at matched parameter budgets, improves calibration (lower ECE/Brier), and reduces inference latency due to geometric locality. These properties make SiFEN a compact, interpretable, and theoretically grounded alternative to dense MLPs and edge-spline networks.
title Simplex-FEM Networks (SiFEN): Learning A Triangulated Function Approximator
topic Machine Learning
url https://arxiv.org/abs/2511.04804