Learning to optimize: A tutorial for continuous and mixed-integer optimization

X Chen, J Liu, W Yin - Science China Mathematics, 2024 - Springer
Learning to optimize (L2O) stands at the intersection of traditional optimization and machine
learning, utilizing the capabilities of machine learning to enhance conventional optimization …

Differentiating through integer linear programs with quadratic regularization and davis-yin splitting

D McKenzie, H Heaton, SW Fung - Transactions on Machine …, 2024 - openreview.net
In many applications, a combinatorial problem must be repeatedly solved with similar, but
distinct parameters. Yet, the parameters $ w $ are not directly observed; only contextual data …

A Mathematics-Inspired Learning-to-Optimize Framework for Decentralized Optimization

Y He, Q Shang, X Huang, J Liu, K Yuan - arXiv preprint arXiv:2410.01700, 2024 - arxiv.org
Most decentralized optimization algorithms are handcrafted. While endowed with strong
theoretical guarantees, these algorithms generally target a broad class of problems, thereby …

A Markovian model for learning-to-optimize

M Sucker, P Ochs - arXiv preprint arXiv:2408.11629, 2024 - arxiv.org
We present a probabilistic model for stochastic iterative algorithms with the use case of
optimization algorithms in mind. Based on this model, we present PAC-Bayesian …

ODE-based Learning to Optimize

Z Xie, W Yin, Z Wen - arXiv preprint arXiv:2406.02006, 2024 - arxiv.org
Recent years have seen a growing interest in understanding acceleration methods through
the lens of ordinary differential equations (ODEs). Despite the theoretical advancements …

From Learning to Optimize to Learning Optimization Algorithms

C Castera, P Ochs - arXiv preprint arXiv:2405.18222, 2024 - arxiv.org
Towards designing learned optimization algorithms that are usable beyond their training
setting, we identify key principles that classical algorithms obey, but have up to now, not …

A Generalization Result for Convergence in Learning-to-Optimize

M Sucker, P Ochs - arXiv preprint arXiv:2410.07704, 2024 - arxiv.org
Convergence in learning-to-optimize is hardly studied, because conventional convergence
guarantees in optimization are based on geometric arguments, which cannot be applied …

Low-Complexity CSI Feedback for FDD Massive MIMO Systems via Learning to Optimize

Y Ma, H He, S Song, J Zhang, KB Letaief - arXiv preprint arXiv:2406.16323, 2024 - arxiv.org
In frequency-division duplex (FDD) massive multiple-input multiple-output (MIMO) systems,
the growing number of base station antennas leads to prohibitive feedback overhead for …

IPM-LSTM: A Learning-Based Interior Point Method for Solving Nonlinear Programs

X Gao, J Xiong, A Wang, Q Duan, J Xue… - arXiv preprint arXiv …, 2024 - arxiv.org
Solving constrained nonlinear programs (NLPs) is of great importance in various domains
such as power systems, robotics, and wireless communication networks. One widely used …

A Learning-only Method for Multi-Cell Multi-User MIMO Sum Rate Maximization

Q Song, J Wang, J Li, G Liu, H Xu - IEEE INFOCOM 2024-IEEE …, 2024 - ieeexplore.ieee.org
Solving the sum rate maximization problem for interference reduction in multi-cell multi-user
multiple-input multiple-output (MIMO) wireless communication systems has been …