mlsanityData to 7 Oct 2026

Accelerating Quadratic Optimization with Reinforcement Learning

arXiv 2107.10847 ↗First posted 22 Jul 2021cs.LG math.OCcited by 30

Not on the map. Why?

Abstract (arXiv)

First-order methods for quadratic optimization such as OSQP are widely used for large-scale machine learning and embedded optimal control, where many related problems must be rapidly solved. These methods face two persistent challenges: manual hyperparameter tuning and convergence time to high-accuracy solutions. To address these, we explore how Reinforcement Learning (RL) can learn a policy to tune parameters to accelerate convergence. In experiments with well-known QP benchmarks we find that our RL policy, RLQP, significantly outperforms state-of-the-art QP solvers by up to 3x. RLQP generalizes surprisingly well to previously unseen problems with varying dimension and structure from different applications, including the QPLIB, Netlib LP and Maros-Meszaros problems. Code for RLQP is available at https://github.com/berkeleyautomation/rlqp.

Key facts. Cited by 30 papers in the map (data through 7 Oct 2026). Builds on 24 of its 50 references that are in the map. First posted on arXiv 22 Jul 2021.

Builds on

24 of 50 references.

References · 24

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  4. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,635
  5. 2015Continuous control with deep reinforcement learningcited by 3,279
  6. 2013Playing Atari with Deep Reinforcement Learningcited by 3,211
  7. 2018Addressing Function Approximation Error in Actor-Critic Methodscited by 2,302
  8. 2017Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environmentscited by 1,519
  9. 2018Relational inductive biases, deep learning, and graph networkscited by 1,341
  10. 2018On First-Order Meta-Learning Algorithmscited by 1,003
  11. 2018IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architecturescited by 866
  12. 2018Machine Learning for Combinatorial Optimization: a Methodological Tour d'Horizoncited by 612
12 more
  1. 2017Learning Combinatorial Optimization Algorithms over Graphscited by 557
  2. 2016Learning to Optimizecited by 195
  3. 2018Learning to Perform Local Rewriting for Combinatorial Optimizationcited by 183
  4. 2017OSQP: An Operator Splitting Solver for Quadratic Programscited by 156
  5. 2020One Policy to Control Them All: Shared Modular Policies for Agent-Agnostic Controlcited by 114
  6. 2018Understanding and correcting pathologies in the training of learned optimizerscited by 97
  7. 2019Checkmate: Breaking the Memory Wall with Optimal Tensor Rematerializationcited by 90
  8. 2018Learning to Branchcited by 78
  9. 2020Tuning-free Plug-and-Play Proximal Algorithm for Inverse Imaging Problemscited by 38
  10. 2018The Voice of Optimizationcited by 28
  11. 2014Linear Convergence and Metric Selection for Douglas-Rachford Splitting and ADMMcited by 4
  12. 2020GOMP: Grasp-Optimized Motion Planning for Bin Pickingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 2; 2022: 6; 2023: 3; 2024: 4; 2025: 8; 2026: 7.

30 papers.

Sort:

By area:

  1. 2022Tutorial on amortized optimizationcited by 83
  2. 2022Automated Dynamic Algorithm Configurationcited by 37
  3. 2023Learning to Warm-Start Fixed-Point Optimization Algorithmscited by 32
  4. 2022A Closer Look at Learned Optimization: Stability, Robustness, and Inductive Biasescited by 30
  5. 2024Expressive Power of Graph Neural Networks for (Mixed-Integer) Quadratic Programscited by 16
  6. 2024Artificial Intelligence for Operations Research: Revolutionizing the Operations Research Processcited by 11
  7. 2025Efficient Action-Constrained Reinforcement Learning via Acceptance-Rejection Method and Augmented MDPscited by 8
  8. 2022Learning context-aware adaptive solvers to accelerate quadratic programmingcited by 7
  9. 2022Optimization for Amortized Inverse Problemscited by 6
  10. 2021A Reinforcement Learning Approach to Parameter Selection for Distributed Optimal Power Flowcited by 5
  11. 2024Data-Driven Performance Guarantees for Classical and Learned Optimizerscited by 5
  12. 2025Towards graph neural networks for provably solving convex optimization problemscited by 4
18 more
  1. 2025Deep FlexQP: Accelerated Nonlinear Programming via Deep Unfoldingcited by 4
  2. 2024NeuralQP: A General Hypergraph-based Optimization Framework for Large-scale QCQPscited by 3
  3. 2025Learning to optimize with guarantees: a complete characterization of linearly convergent algorithmscited by 3
  4. 2025Learning to accelerate distributed ADMM using graph neural networkscited by 3
  5. 2023Hint assisted reinforcement learning: an application in radio astronomycited by 2
  6. 2025Warm-starting active-set solvers using graph neural networkscited by 2
  7. 2026Learning to accelerate Krasnosel'skii-Mann fixed-point iterations with guaranteescited by 2
  8. 2026Scalable Deep Unfolding of Conic Optimizerscited by 2
  9. 2021NICE: Robust Scheduling through Reinforcement Learning-Guided Integer Programmingcited by 1
  10. 2026Parametric Nonconvex Optimization via Convex Surrogatescited by 1
  11. 2022Bridging the gap between QP-based and MPC-based RLcited by 0
  12. 2023Multi-Abstractive Neural Controller: An Efficient Hierarchical Control Architecture for Interactive Drivingcited by 0
  13. 2025Reinforcement learning for adaptive interior point methods in convex quadratic programmingcited by 0
  14. 2025Data-driven Projection Generation for Efficiently Solving Heterogeneous Quadratic Programming Problemscited by 0
  15. 2026Learning Over-Relaxation Policies for ADMM with Convergence Guaranteescited by 0
  16. 2026When Do Learned Diffusion Proposals Help Constraint Solving? A Controlled Study on Continuous Algebraic Systemscited by 0
  17. 2026Reinforcement Learning in Operational Research: A Technical Review and Practical Roadmapcited by 0
  18. 2026Reinforcement Learning to Accelerate Primal-Dual Hybrid Gradient for Linear Programmingcited by 0