Open Peer Review. Open Publishing. Open Access. Open Discussion. Open Directory. Open Recommendations. Open API. Open Source.
On the insufficiency of existing momentum schemes for Stochastic Optimization
Rahul Kidambi, Praneeth Netrapalli, Prateek Jain, Sham M. Kakade
Feb 15, 2018 (modified: Feb 15, 2018)ICLR 2018 Conference Blind Submissionreaders: everyoneShow Bibtex
Abstract:Momentum based stochastic gradient methods such as heavy ball (HB) and Nesterov's accelerated gradient descent (NAG) method are widely used in practice for training deep networks and other supervised learning models, as they often provide significant improvements over stochastic gradient descent (SGD). Theoretically, these ```fast gradient methods have provable improvements over gradient descent only for the deterministic case, where the gradients are exact. In the stochastic case, the popular explanations for their wide applicability is that when these fast gradient methods are applied in the stochastic case, they partially mimic their exact gradient counterparts, resulting in some practical gain. This work provides a counterpoint to this belief by proving that there are simple problem instances where these methods cannot outperform SGD despite the best setting of its parameters. These negative problem instances are, in an informal sense, generic; they do not look like carefully constructed pathological instances. These results suggest (along with empirical evidence) that HB or NAG's practical performance gains are a by-product of minibatching.
Furthermore, this work provides a viable (and provable) alternative, which, on the same set of problem instances, significantly improves over HB, NAG, and SGD's performance. This algorithm, referred to as Accelerated Stochastic Gradient Descent (ASGD), is a simple to implement stochastic algorithm, based on a relatively less popular variant of Nesterov's Accelerated Gradient Descent. Extensive empirical results in this paper show that ASGD has performance gains over HB, NAG, and SGD.
TL;DR:Existing momentum/acceleration schemes such as heavy ball method and Nesterov's acceleration employed with stochastic gradients do not improve over vanilla stochastic gradient descent, especially when employed with small batch sizes.
Keywords:Stochastic Gradient Descent, Deep Learning, Momentum, Acceleration, Heavy Ball, Nesterov Acceleration, Stochastic Optimization, SGD, Accelerated Stochastic Gradient Descent
Enter your feedback below and we'll get back to you as soon as possible.