arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1311.2115cs.LG
收藏

统一随机梯度法和拟牛顿法的快速大规模优化

Fast large-scale optimization by unifying stochastic gradient and quasi-Newton methods

Jascha Sohl-Dickstein, Ben Poole, Surya Ganguli

摘要:

我们提出了一种算法,用于最小化的功能,结合了计算效率的随机梯度下降(SGD)的二阶曲率信息利用拟牛顿方法的总和。我们统一这些不同的方法,保持一个独立的海森近似的每个贡献函数的总和。我们保持计算的易处理性和限制内存的要求,即使是高维的优化问题,通过存储和操纵这些二次近似在一个共享的,随时间变化的,低维的子空间。每个更新步骤只需要一个贡献函数或小批量评估(如SGD),每个步骤都使用近似逆Hessian进行缩放,并且几乎不需要调整超参数(如拟牛顿方法的典型)。该算法与早期的随机二阶技术相比,该技术将每个贡献函数的Hessian视为对完整Hessian的噪声近似,而不是作为直接估计的目标。我们的实验证明了七个不同的优化问题的收敛性。该算法作为开源Python和MATLAB软件包发布。

英文摘要:

We present an algorithm for minimizing a sum of functions that combines the computational efficiency of stochastic gradient descent (SGD) with the second order curvature information leveraged by quasi-Newton methods. We unify these disparate approaches by maintaining an independent Hessian approximation for each contributing function in the sum. We maintain computational tractability and limit memory requirements even for high dimensional optimization problems by storing and manipulating these quadratic approximations in a shared, time evolving, low dimensional subspace. Each update step requires only a single contributing function or minibatch evaluation (as in SGD), and each step is scaled using an approximate inverse Hessian and little to no adjustment of hyperparameters is required (as is typical for quasi-Newton methods). This algorithm contrasts with earlier stochastic second order techniques that treat the Hessian of each contributing function as a noisy approximation to the full Hessian, rather than as a target for direct estimation. We experimentally demonstrate improved convergence on seven diverse optimization problems. The algorithm is released as open source Python and MATLAB packages.

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。

↑