Exact Convex Reformulations of Linear Neural Networks via Completely Positive Lifting

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Prakhya, Karthik, Yurtsever, Alp
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918508202295296
author Prakhya, Karthik
Yurtsever, Alp
author_facet Prakhya, Karthik
Yurtsever, Alp
contents We show that the training problem of a deep linear neural network under the squared loss admits an exact convex reformulation in a lifted space over a generalized completely positive cone. The reformulation has the same optimal value as the original nonconvex problem and is linear in the lifted variables, with all nonconvexity encoded in the cone constraint. Its ambient lifted dimension depends only on the input and output dimensions, independent of the network depth and the number of data points, and the bottleneck width enters only through scalar constraints. The construction proceeds by reducing the multilayer parameterization to a bilinear factorization, lifting it to a rank-constrained semidefinite program, expressing the rank constraint via a complementarity condition, and applying a completely positive lifting. While the resulting formulation is computationally intractable in general, it gives an exact conic representation of the nonconvexity induced by linear factorization and connects linear neural network training with copositive programming.
format Preprint
id arxiv_https___arxiv_org_abs_2605_17692
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Exact Convex Reformulations of Linear Neural Networks via Completely Positive Lifting
Prakhya, Karthik
Yurtsever, Alp
Machine Learning
Optimization and Control
90C25, 90C22, 90C26
We show that the training problem of a deep linear neural network under the squared loss admits an exact convex reformulation in a lifted space over a generalized completely positive cone. The reformulation has the same optimal value as the original nonconvex problem and is linear in the lifted variables, with all nonconvexity encoded in the cone constraint. Its ambient lifted dimension depends only on the input and output dimensions, independent of the network depth and the number of data points, and the bottleneck width enters only through scalar constraints. The construction proceeds by reducing the multilayer parameterization to a bilinear factorization, lifting it to a rank-constrained semidefinite program, expressing the rank constraint via a complementarity condition, and applying a completely positive lifting. While the resulting formulation is computationally intractable in general, it gives an exact conic representation of the nonconvexity induced by linear factorization and connects linear neural network training with copositive programming.
title Exact Convex Reformulations of Linear Neural Networks via Completely Positive Lifting
topic Machine Learning
Optimization and Control
90C25, 90C22, 90C26
url https://arxiv.org/abs/2605.17692