Proposes a method to improve Byzantine-robustness in compressed federated learning.
problem Byzantine-robustness in compressed federated learning.
method Gradient difference compression and stochastic average gradient algorithm (SAGA).
result The proposed method reaches a neighborhood of the optimal solution at a linear convergence rate.
BICompFL tackles bi-directional compression challenges in stochastic FL, reducing communication costs by an order of magnitude.
problem Communication bottleneck in federated learning, especially with stochastic updates.
method Introduces BICompFL, a bi-directional compression approach for stochastic federated learning.
result Significantly reduces communication costs (by an order of magnitude) while maintaining accuracy.
A new framework compresses neural networks using sparse optimization.
problem Efficiently reducing the size of deep neural networks for practical deployment.
method Sparse optimization for model compression, tailored for stochastic learning.
result Up to 7.2 and 2.9 times FLOPs reduction with comparable accuracy.
Paper proposes SCALLION and SCAFCOM for compressed FL with reduced communication.
problem Reducing communication overhead in Federated Learning with data heterogeneity and partial participation.
method Revisit and simplify stochastic controlled averaging, proposing SCALLION and SCAFCOM for unbiased and biased compression.
result SCALLION and SCAFCOM outperform existing methods in communication and computation complexities.
Paper improves compressed SGD to reach second-order stationary points.
problem Efficiently reaching second-order stationary points in distributed machine learning.
method Gradient compression and RandomK compressor to improve convergence to second-order stationary points.
result Compressed SGD reaches second-order stationary points with improved communication efficiency.
New method improves image compression using bits-back coding.
problem Lossy image compression with deep latent variable models.
method Iterative inference, stochastic annealing, bits-back coding.
result New state-of-the-art performance on lossy image compression.
New biased compression methods lead to faster convergence in distributed learning.
problem Improving convergence rates in distributed learning with biased compression.
method Study of three classes of biased compression operators in distributed learning.
result Biased compressors can lead to linear convergence rates in both single node and distributed settings.
New SGD variant makes neural networks compressible without assumptions.
problem Improving neural network compressibility without strong assumptions.
method Introducing heavy-tailed noise to SGD iterates.
result Compressible outputs with high probability for any compression rate.
Data compression is a popular technique for improving the efficiency of data processing workloads such as SQL queries and more recently, machine learning (ML) with classical batch gradient methods. But the efficacy of such ideas for mini-batch stochastic gradient descent (MGD), arguably the workhorse algorithm of moder…
Paper analyzes convergence rates of compressed LSR algorithms in federated learning.
problem Impact of compression on convergence rates in distributed learning.
method Analyzes a general stochastic approximation algorithm for LSR with weak assumptions.
result Convergence rates depend on the covariance of additive noise and compression strategy.
New hierarchical VQ-VAE scheme improves image compression quality and features at low bitrates.
problem Low bitrate image compression maintaining quality and features.
method Hierarchical VQ-VAE with stochastic quantization and Markovian latent variables.
result High perceptual quality and semantic features at low bitrates.
IntSGD compresses SGD gradients without floats, converging as SGD.
problem Efficiently compressing stochastic gradients in distributed SGD.
method Adaptive integer compression of gradients, estimating scaling adaptively.
result IntSGD matches SGD's iteration complexity for convex and non-convex functions.
Paper proposes a compression principle for neural networks using Bayesian optimization.
problem Finding methods for making generalizable predictions in machine learning.
method Compression principle and Bayesian optimization approach.
result Optimal predictive models minimize total compressed message length of data and model definition.
New method reduces communication costs in distributed nonconvex optimization.
problem Large communication costs between central server and local workers in distributed learning.
method Communication-compressed AMSGrad for distributed nonconvex optimization.
result Converges to first-order stationary point with same iteration complexity as vanilla AMSGrad.
New method reduces FL communication cost by leveraging server's side information.
problem High communication cost in federated learning.
method Exploits server's side information to compress client updates.
result Up to 82 times smaller bitrate with higher accuracy.
The online problem of computing the top eigenvector is fundamental to machine learning. In both adversarial and stochastic settings, previous results (such as matrix multiplicative weight update, follow the regularized leader, follow the compressed leader, block power method) either achieve optimal regret but run slow,…
Unified sign-based compression for federated learning with faster convergence.
problem High communication cost in federated learning with large-scale models.
method Unified noisy perturbation scheme for sign-based compression.
result Achieves faster convergence rate than existing sign-based methods.
This work proposes ACTC for adaptive distributed learning under communication constraints.
problem Adaptive distributed learning in networks with communication constraints.
method ACTC (Adapt-Compress-Then-Combine) strategy with diffusion exchange of compressed updates.
result ACTC iterates converge to the optimizer with significant bit savings.
We propose and analyze a new type of stochastic first order method: gradient descent with compressed iterates (GDCI). GDCI in each iteration first compresses the current iterate using a lossy randomized compression technique, and subsequently takes a gradient step. This method is a distillation of a key ingredient in t…
Unified framework for distributed compressed SGD under (L0,L1)-smoothness.
problem Understanding the joint effect of batch noise, adaptivity, and compression in distributed stochastic optimization.
method Developed a unified theoretical framework using SDEs that incorporate curvature-dependent terms.
result Normalizing updates in DCSGD stabilizes convergence, with normalization degree determined by noise structure and landscape regularity.
We consider decentralized stochastic optimization with the objective function (e.g. data samples for machine learning task) being distributed over n machines that can only communicate to their neighbors on a fixed communication graph. To reduce the communication bottleneck, the nodes compress (e.g. quantize or sparsi…
Unified bounds linking compressibility, fractal dimensions, and mutual information.
problem Understanding generalization in stochastic learning algorithms.
method Rate-distortion theory applied to machine learning generalization.
result Unified bounds linking compressibility, fractal dimensions, and mutual information.
Paper introduces MoTEF for faster decentralized optimization with compressed communication.
problem Efficiency bottleneck in decentralized machine learning applications.
method Integrates communication compression with Momentum Tracking and Error Feedback.
result Significantly outperforms existing methods under arbitrary data heterogeneity.
Improved neural image compression with refined latent representations.
problem Sub-optimal results from variational autoencoders due to imperfect optimization and capacity limitations.
method Stochastic Gumbel Annealing (SGA) and its extensions (SGA+), including three different methods.
result Significant improvement in compression performance, especially on the R-D trade-off.
Deep neural networks have dramatically achieved great success on a variety of challenging tasks. However, most successful DNNs have an extremely complex structure, leading to extensive research on model compression.As a significant area of progress in model compression, traditional gradual pruning approaches involve an…
Recurrent iterated function systems (RIFSs) are improvements of iterated function systems (IFSs) using elements of the theory of Marcovian stochastic processes which can produce more natural looking images. We construct new RIFSs consisting substantially of a vertical contraction factor function and nonlinear transform…
Gradient sparsification enhances privacy-preserving machine learning models.
problem Improving performance of differentially-private machine learning models under privacy constraints.
method Gradient sparsification combined with compressed sensing and additive Laplace noise.
result Gradient sparsification can improve performance of differentially-private machine learning models for small privacy budgets.
Proposes a new method to selectively access privileged information in reinforcement learning.
problem Selective compression of privileged information in reinforcement learning.
method Formulates a variational bandwidth bottleneck to decide stochastically whether to access privileged information.
result Improves generalization and reduces access to costly information in reinforcement learning experiments.
We solve the compressive sensing problem via convolutional factor analysis, where the convolutional dictionaries are learned {\em in situ} from the compressed measurements. An alternating direction method of multipliers (ADMM) paradigm for compressive sensing inversion based on convolutional factor analysis is develope…
Deep Gaussian processes provide a flexible approach to probabilistic modelling of data using either supervised or unsupervised learning. For tractable inference approximations to the marginal likelihood of the model must be made. The original approach to approximate inference in these models used variational compressio…
NeuralChaos efficiently approximates complex stochastic processes.
problem Representing and computing square-integrable predictable processes over time.
method Introduces NeuralChaos, a neural operator architecture for Rd-valued predictable processes. result NeuralChaos achieves best N-term chaoslet approximation rates and is dense in HT2(Rd). New coding theorem shows achievable rate matches theoretical limit.
problem Unknown existence of encoders and decoders for RDPF.
method Used stochastic, variable-length codes to prove RDPF achievable.
result Achievable rate matches theoretical rate-distortion-perception function.
Artemis framework improves distributed learning with bidirectional compression and partial participation.
problem Learning in distributed or federated settings with communication constraints and device partial participation.
method Artemis framework using bidirectional compression, memory mechanism, and Polyak-Ruppert averaging.
result Fast rates of convergence (linear up to a threshold) under weak assumptions on stochastic gradients.
Quantum computing speeds up analysis of financial stochastic processes.
problem Challenging simulation and analysis of continuous time stochastic processes.
method Established a quantum framework for efficient state preparation and information extraction.
result Extraction of path-dependent and history-sensitive information from stochastic processes efficiently.
New study reveals how heavy-tailed SGD dynamics lead to compressible neural networks.
problem Understanding why large neural networks can be compressed effectively.
method Linking SGD dynamics to compressibility properties of neural networks.
result Large step-size/batch-size ratios and overparametrization lead to heavy-tailed SGD dynamics, making networks compressible.
New algorithm reduces dimensionality in stochastic optimization.
problem Stochastic optimization in high-dimensional problems.
method Proposes a sparsity-inducing stochastic gradient-free (SI-SGF) algorithm.
result Proves dimension-free query complexity in convex and strongly convex cases.
Adaptive step-size method improves compressed SGD performance in machine learning.
problem Communication bottleneck in distributed and decentralized optimization.
method Developed an adaptive step-size method for compressed SGD.
result Order-optimal convergence rates for various objective functions.
Communication is a key bottleneck in distributed training. Recently, an \emph{error-compensated} compression technology was particularly designed for the \emph{centralized} learning and receives huge successes, by showing significant advantages over state-of-the-art compression based methods in saving the communication…
We analyze (stochastic) gradient descent (SGD) with delayed updates on smooth quasi-convex and non-convex functions and derive concise, non-asymptotic, convergence rates. We show that the rate of convergence in all cases consists of two terms: (i) a stochastic term which is not affected by the delay, and (ii) a higher …
A new algorithm improves Bayesian federated learning by reducing communication overhead.
problem Bayesian federated learning constraints, including privacy, data ownership, and communication overhead.
method Proposes Quantised Langevin Stochastic Dynamics (QLSD) for Bayesian federated learning, using gradient compression and variance reduction techniques.
result Non-asymptotic and asymptotic convergence guarantees for QLSD and its improved versions.
Paper uses SGLD to recover signals from generative models, proving convergence under mild conditions.
problem Signal recovery from generative priors in compressed sensing.
method Stochastic Gradient Langevin Dynamics (SGLD) for signal recovery.
result SGLD converges to the true signal under mild assumptions on the generative model.
Improved online Sinkhorn algorithm for large-scale data processing.
problem Efficiency of Sinkhorn algorithm for large-scale data streams.
method Revisit and improve convergence analysis of online Sinkhorn algorithm, introduce compressed online Sinkhorn algorithm.
result New faster convergence rate for online Sinkhorn algorithm under certain conditions.
Unified analysis of SGD variants for nonconvex federated optimization.
problem Performance of stochastic gradient methods in nonconvex optimization.
method Proposed a unified assumption for modeling stochastic gradient second moment, leading to a single convergence analysis for various methods.
result Unified convergence analysis for a wide range of SGD variants and distributed methods.
Sketching reduces data size for accurate spectral estimation.
problem Estimating spectral density from large simulation datasets.
method Sketching for dimensionality reduction and data compression.
result Sketching provides 90% accurate spectral density estimate with 10% data.
Large-scale machine learning models are often trained by parallel stochastic gradient descent algorithms. However, the communication cost of gradient aggregation and model synchronization between the master and worker nodes becomes the major obstacle for efficient learning as the number of workers and the dimension of …
New method reduces communication in deep learning training.
problem Communication overhead in distributed deep learning training.
method Random-block sparsification to reduce gradients communicated.
result Performance close to standard SGD with reduced communication.
Optimizing distributed learning systems is an art of balancing between computation and communication. There have been two lines of research that try to deal with slower networks: {\em communication compression} for low bandwidth networks, and {\em decentralization} for high latency networks. In this paper, We explore a…
Stochastic encoders outperform deterministic ones in 'perfect perceptual quality'.
problem Understanding when stochastic encoders outperform deterministic ones.
method Provided a toy example to illustrate performance.
result Stochastic encoders can significantly outperform deterministic ones in 'perfect perceptual quality'.