ترغب بنشر مسار تعليمي؟ اضغط هنا

Almost Tune-Free Variance Reduction

172   0   0.0 ( 0 )
 نشر من قبل Bingcong Li
 تاريخ النشر 2019
  مجال البحث الهندسة المعلوماتية
والبحث باللغة English




اسأل ChatGPT حول البحث

The variance reduction class of algorithms including the representative ones, SVRG and SARAH, have well documented merits for empirical risk minimization problems. However, they require grid search to tune parameters (step size and the number of iterations per inner loop) for optimal performance. This work introduces `almost tune-free SVRG and SARAH schemes equipped with i) Barzilai-Borwein (BB) step sizes; ii) averaging; and, iii) the inner loop length adjusted to the BB step sizes. In particular, SVRG, SARAH, and their BB variants are first reexamined through an `estimate sequence lens to enable new averaging methods that tighten their convergence rates theoretically, and improve their performance empirically when the step size or the inner loop length is chosen large. Then a simple yet effective means to adjust the number of iterations per inner loop is developed to enhance the merits of the proposed averaging schemes and BB step sizes. Numerical tests corroborate the proposed methods.



قيم البحث

اقرأ أيضاً

Variance reduction (VR) methods for finite-sum minimization typically require the knowledge of problem-dependent constants that are often unknown and difficult to estimate. To address this, we use ideas from adaptive gradient methods to propose AdaSV RG, which is a fully adaptive variant of SVRG, a common VR method. AdaSVRG uses AdaGrad in the inner loop of SVRG, making it robust to the choice of step-size, and allowing it to adaptively determine the length of each inner-loop. When minimizing a sum of $n$ smooth convex functions, we prove that AdaSVRG requires $O(n + 1/epsilon)$ gradient evaluations to achieve an $epsilon$-suboptimality, matching the typical rate, but without needing to know problem-dependent constants. However, VR methods including AdaSVRG are slower than SGD when used with over-parameterized models capable of interpolating the training data. Hence, we also propose a hybrid algorithm that can adaptively switch from AdaGrad to AdaSVRG, achieving the best of both stochastic gradient and VR methods, but without needing to tune their step-sizes. Via experiments on synthetic and standard real-world datasets, we validate the robustness and effectiveness of AdaSVRG, demonstrating its superior performance over other tune-free VR methods.
141 - Rong Ge , Zhize Li , Weiyao Wang 2019
Variance reduction techniques like SVRG provide simple and fast algorithms for optimizing a convex finite-sum objective. For nonconvex objectives, these techniques can also find a first-order stationary point (with small gradient). However, in noncon vex optimization it is often crucial to find a second-order stationary point (with small gradient and almost PSD hessian). In this paper, we show that Stabilized SVRG (a simple variant of SVRG) can find an $epsilon$-second-order stationary point using only $widetilde{O}(n^{2/3}/epsilon^2+n/epsilon^{1.5})$ stochastic gradients. To our best knowledge, this is the first second-order guarantee for a simple variant of SVRG. The running time almost matches the known guarantees for finding $epsilon$-first-order stationary points.
Stochastic gradient Langevin dynamics (SGLD) has gained the attention of optimization researchers due to its global optimization properties. This paper proves an improved convergence property to local minimizers of nonconvex objective functions using SGLD accelerated by variance reductions. Moreover, we prove an ergodicity property of the SGLD scheme, which gives insights on its potential to find global minimizers of nonconvex objectives.
Alternating Direction Method of Multipliers (ADMM) is a popular method in solving Machine Learning problems. Stochastic ADMM was firstly proposed in order to reduce the per iteration computational complexity, which is more suitable for big data probl ems. Recently, variance reduction techniques have been integrated with stochastic ADMM in order to get a fast convergence rate, such as SAG-ADMM and SVRG-ADMM,but the convergence is still suboptimal w.r.t the smoothness constant. In this paper, we propose a new accelerated stochastic ADMM algorithm with variance reduction, which enjoys a faster convergence than all the other stochastic ADMM algorithms. We theoretically analyze its convergence rate and show its dependence on the smoothness constant is optimal. We also empirically validate its effectiveness and show its priority over other stochastic ADMM algorithms.
Effective hyper-parameter tuning is essential to guarantee the performance that neural networks have come to be known for. In this work, a principled approach to choosing the learning rate is proposed for shallow feedforward neural networks. We assoc iate the learning rate with the gradient Lipschitz constant of the objective to be minimized while training. An upper bound on the mentioned constant is derived and a search algorithm, which always results in non-divergent traces, is proposed to exploit the derived bound. It is shown through simulations that the proposed search method significantly outperforms the existing tuning methods such as Tree Parzen Estimators (TPE). The proposed method is applied to three different existing applications: a) channel estimation in OFDM systems, b) prediction of the exchange currency rates and c) offset estimation in OFDM receivers, and it is shown to pick better learning rates than the existing methods using the same or lesser compute power.

الأسئلة المقترحة

التعليقات
جاري جلب التعليقات جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا