This content will become publicly available on November 29, 2023
 Editors:
 Oh, Alice H.; Agarwal, Alekh; Belgrave, Danielle; Cho, Kyunghyun
 Publication Date:
 NSFPAR ID:
 10389696
 Journal Name:
 Advances in neural information processing systems
 ISSN:
 10495258
 Sponsoring Org:
 National Science Foundation
More Like this

Providing privacy protection has been one of the primary motivations of Federated Learning (FL). Recently, there has been a line of work on incorporating the formal privacy notion of differential privacy with FL. To guarantee the clientlevel differential privacy in FL algorithms, the clients’ transmitted model updates have to be clipped before adding privacy noise. Such clipping operation is substantially different from its counterpart of gradient clipping in the centralized differentially private SGD and has not been wellunderstood. In this paper, we first empirically demonstrate that the clipped FedAvg can perform surprisingly well even with substantial data heterogeneity when training neural networks, which is partly because the clients’ updates become similar for several popular deep architectures. Based on this key observation, we provide the convergence analysis of a differential private (DP) FedAvg algorithm and highlight the relationship between clipping bias and the distribution of the clients’ updates. To the best of our knowledge, this is the first work that rigorously investigates theoretical and empirical issues regarding the clipping operation in FL algorithms.

Stochastic gradient descent with momentum (SGDm) is one of the most popular optimization algorithms in deep learning. While there is a rich theory of SGDm for convex problems, the theory is considerably less developed in the context of deep learning where the problem is nonconvex and the gradient noise might exhibit a heavytailed behavior, as empirically observed in recent studies. In this study, we consider a \emph{continuoustime} variant of SGDm, known as the underdamped Langevin dynamics (ULD), and investigate its asymptotic properties under heavytailed perturbations. Supported by recent studies from statistical physics, we argue both theoretically and empirically that the heavytails of such perturbations can result in a bias even when the stepsize is small, in the sense that \emph{the optima of stationary distribution} of the dynamics might not match \emph{the optima of the cost function to be optimized}. As a remedy, we develop a novel framework, which we coin as \emph{fractional} ULD (FULD), and prove that FULD targets the socalled Gibbs distribution, whose optima exactly match the optima of the original cost. We observe that the Euler discretization of FULD has noteworthy algorithmic similarities with \emph{natural gradient} methods and \emph{gradient clipping}, bringing a new perspective on understanding their rolemore »

Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)The Neuronix highperformance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from lowend consumer grade devices such as the Nvidia GTX 970 to higherend devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely computeintensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floatingpoint calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should producemore »

Deep learning architectures are usually proposed with millions of parameters, resulting in a memory issue when training deep neural networks with stochastic gradient descent type methods using large batch sizes. However, training with small batch sizes tends to produce low quality solution due to the large variance of stochastic gradients. In this paper, we tackle this problem by proposing a new framework for training deep neural network with small batches/noisy gradient. During optimization, our method iteratively applies a proximal type regularizer to make loss function strongly convex. Such regularizer stablizes the gradient, leading to better training performance. We prove that our algorithm achieves comparable convergence rate as vanilla SGD even with small batch size. Our framework is simple to implement and can be potentially combined with many existing optimization algorithms. Empirical results show that our method outperforms SGD and Adam when batch size is small. Our implementation is available at https://github.com/huiqu18/TRAlgorithm.

Despite the existence of divergence examples, RMSprop remains one of the most popular algorithms in machine learning. Towards closing the gap between theory and practice, we prove that RMSprop converges with proper choice of hyperparameters under certain conditions. More specifically, we prove that when the hyperparameter beta_2 is close enough to 1, RMSprop and its random shuffling version converge to a bounded region in general, and to critical points in the interpolation regime. It is worth mentioning that our results do not depend on “bounded gradient" assumption, which is often the key assumption utilized by existing theoretical work for Adamtype adaptive gradient method. Removing this assumption allows us to establish a phase transition from divergence to nondivergence for RMSprop. Finally, based on our theory, we conjecture that in practice there is a critical threshold \beta_2 , such that RMSprop generates reasonably good results only if 1 > \beta_2\ge \beta^*_2 . We provide empirical evidence for such a phase transition in our numerical experiments.