Pith. sign in

REVIEW 5 cited by

Learning Constrained Optimization with Deep Augmented Lagrangian Methods

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2403.03454 v2 pith:MJG73SD3 submitted 2024-03-06 cs.LG math.OC

classification cs.LGmath.OC
keywords learningdualconstrainedestimatesmethodsoptimizationprimalscheme
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Learning to Optimize (LtO) is a problem setting in which a machine learning (ML) model is trained to emulate a constrained optimization solver. Learning to produce optimal and feasible solutions subject to complex constraints is a difficult task, but is often made possible by restricting the input space to a limited distribution of related problems. Most LtO methods focus on directly learning solutions to the primal problem, and applying correction schemes or loss function penalties to encourage feasibility. This paper proposes an alternative approach, in which the ML model is trained instead to predict dual solution estimates directly, from which primal estimates are constructed to form dual-feasible solution pairs. This enables an end-to-end training scheme is which the dual objective is maximized as a loss function, and solution estimates iterate toward primal feasibility, emulating a Dual Ascent method. First it is shown that the poor convergence properties of classical Dual Ascent are reflected in poor convergence of the proposed training scheme. Then, by incorporating techniques from practical Augmented Lagrangian methods, we show how the training scheme can be improved to learn highly accurate constrained optimization solvers, for both convex and nonconvex problems.

Discussion (0). Sign in to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Training with Hard Constraints: Learning Neural Certificates and Controllers for SDEs

    eess.SY 2026-02 conditional novelty 6.0 of 10

    Neural reach-avoid certificates for SDEs can be trained with hard guarantees via a bound-based loss, or with PAC guarantees via scenario optimization on the last layer.

  2. PGLearn -- An Open-Source Learning Toolkit for Optimal Power Flow

    cs.LG 2025-05 conditional novelty 6.0 of 10

    PGLearn provides a large open-source dataset collection and toolkit with AC, DC, and SOC-OPF primal and dual solutions, time-series data for large grids, and benchmarking tools for ML-based OPF methods.

  3. Large-scale portfolio optimization with variational neural annealing

    cond-mat.dis-nn 2025-07 reject novelty 5.0 of 10

    VNA produces Sharpe-ratio-competitive portfolios on indices up to 2,008 assets, but the claimed speed advantage and universal finite-size scaling are not robustly supported.

  4. Deep Uzawa for Kinetic Transport with Lagrange-Enforced Boundaries

    math.NA 2025-07 conditional novelty 4.0 of 10

    A Lagrange-multiplier Uzawa iteration with neural-network trial functions is developed for stationary transport, with continuum convergence proofs and five qualitative numerical examples.

  5. Learning to Optimize by Differentiable Programming

    cs.MS 2026-01 unverdicted novelty 2.0 of 10

    A tutorial survey of differentiable-programming-based first-order optimization, with dual-based PyTorch case studies and no new results.

Pith tools