arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.17823math.OCcs.LG
收藏

最优的无Hessian/Jacobian非凸-PL双层优化

Optimal Hessian/Jacobian-Free Nonconvex-PL Bilevel Optimization

College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics(南京航空航天大学计算机科学与技术学院)

Feihu Huang

AI总结:

针对非凸-PL双层优化问题计算复杂度高的问题,提出一种无Hessian/Jacobian的高效方法HJFBiO,在温和条件下证明了其最优收敛速率和梯度复杂度,并通过实验验证了有效性。

摘要:

双层优化广泛应用于超参数学习、元学习和强化学习等许多机器学习任务中。尽管最近开发了许多算法来解决双层优化问题,但它们通常依赖于(强)凸的下层问题。最近,一些方法被提出用于解决非凸-PL双层优化问题,其中上层问题可能是非凸的,下层问题也可能是非凸的,同时满足Polyak-Łojasiewicz(PL)条件。然而,这些方法仍然具有较高的收敛复杂度或较高的计算复杂度,例如需要计算昂贵的Hessian/Jacobian矩阵及其逆矩阵。因此,在本文中,我们提出了一种高效的无Hessian/Jacobian方法(即HJFBiO),具有最优的收敛复杂度,以解决非凸-PL双层问题。在理论上,在一些温和的条件下,我们证明了我们的HJFBiO方法获得了$O(\ rac{1}{T})$的最优收敛速率,其中$T$表示迭代次数,并且在寻找$ε$-稳定解时具有$O(ε^{-1})$的最优梯度复杂度。我们在双层PL博弈和超表示学习任务上进行了数值实验,以证明我们提出的方法的有效性。

英文摘要:

Bilevel optimization is widely applied in many machine learning tasks such as hyper-parameter learning, meta learning and reinforcement learning. Although many algorithms recently have been developed to solve the bilevel optimization problems, they generally rely on the (strongly) convex lower-level problems. More recently, some methods have been proposed to solve the nonconvex-PL bilevel optimization problems, where their upper-level problems are possibly nonconvex, and their lower-level problems are also possibly nonconvex while satisfying Polyak-Łojasiewicz (PL) condition. However, these methods still have a high convergence complexity or a high computation complexity such as requiring compute expensive Hessian/Jacobian matrices and its inverses. In the paper, thus, we propose an efficient Hessian/Jacobian-free method (i.e., HJFBiO) with the optimal convergence complexity to solve the nonconvex-PL bilevel problems. Theoretically, under some mild conditions, we prove that our HJFBiO method obtains an optimal convergence rate of $O(\frac{1}{T})$, where $T$ denotes the number of iterations, and has an optimal gradient complexity of $O(ε^{-1})$ in finding an $ε$-stationary solution. We conduct some numerical experiments on the bilevel PL game and hyper-representation learning task to demonstrate efficiency of our proposed method.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑