arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2012.15477stat.MLcs.LG
收藏

粒子对偶平均:平均场神经网络的全局收敛率优化

Particle Dual Averaging: Optimization of Mean Field Neural Networks with Global Convergence Rate Analysis

Kyushu Institute of Technology(九州工业大学);RIKEN Center for Advanced Intelligence Project(理化学研究所革新智能综合研究中心);University of Toronto(多伦多大学);Vector Institute for Artificial Intelligence(矢量人工智能研究院);The University of Tokyo(东京大学)

Atsushi Nitanda, Denny Wu, Taiji Suzuki

摘要:

提出了粒子对偶平均(PDA)方法,将凸优化中的对偶平均方法推广到具有定量运行时间保证的概率分布上的优化.该算法由内循环和外循环组成:内循环利用Langevin算法近似求解一个平稳分布,然后在外循环中进行优化。因此,该方法可以被解释为一个扩展的Langevin算法,自然地处理非线性功能的概率空间。所提出的方法的一个重要应用是在平均场区域中优化神经网络,由于非线性特征学习的存在,这在理论上是有吸引力的,但定量收敛速度可能是具有挑战性的。通过将有限维凸优化理论引入测度空间,分析了正则化经验/期望风险最小化下的PDA,并在更一般的情况下建立了双层平均场神经网络学习的定量全局收敛性.我们的理论结果得到了合理规模的神经网络的数值模拟的支持。

英文摘要:

We propose the particle dual averaging (PDA) method, which generalizes the dual averaging method in convex optimization to the optimization over probability distributions with quantitative runtime guarantee. The algorithm consists of an inner loop and outer loop: the inner loop utilizes the Langevin algorithm to approximately solve for a stationary distribution, which is then optimized in the outer loop. The method can thus be interpreted as an extension of the Langevin algorithm to naturally handle nonlinear functional on the probability space. An important application of the proposed method is the optimization of neural network in the mean field regime, which is theoretically attractive due to the presence of nonlinear feature learning, but quantitative convergence rate can be challenging to obtain. By adapting finite-dimensional convex optimization theory into the space of measures, we analyze PDA in regularized empirical / expected risk minimization, and establish quantitative global convergence in learning two-layer mean field neural networks under more general settings. Our theoretical results are supported by numerical simulations on neural networks with reasonable size.

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。

↑