Accelerating Quadratic Optimization with Reinforcement Learning
This work addresses efficiency issues in quadratic optimization for machine learning and control applications, offering a novel acceleration method.
The paper tackles the challenges of manual hyperparameter tuning and slow convergence in quadratic optimization solvers by using reinforcement learning to learn a tuning policy, resulting in up to 3x speedup over state-of-the-art solvers on benchmarks.
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.