This paper improves Minibatch SGD convergence through typicality sampling.
problem Slow convergence of Minibatch SGD due to large gradient noise.
method Typicality sampling for more efficient batch selection.
result Typical batch SGD outperforms conventional Minibatch SGD in convergence.
Recent work has argued that stochastic gradient descent can approximate the Bayesian uncertainty in model parameters near local minima. In this work we develop a similar correspondence for minibatch natural gradient descent (NGD). We prove that for sufficiently small learning rates, if the model predictions on the trai…
It is well known that, for most datasets, the use of large-size minibatches for Stochastic Gradient Descent (SGD) typically leads to slow convergence and poor generalization. On the other hand, large minibatches are of great practical interest as they allow for a better exploitation of modern GPUs. Previous literature …
Stochastic Gradient Descent (SGD) is a popular optimization method which has been applied to many important machine learning tasks such as Support Vector Machines and Deep Neural Networks. In order to parallelize SGD, minibatch training is often employed. The standard approach is to uniformly sample a minibatch at each…
Minibatching is a very well studied and highly popular technique in supervised learning, used by practitioners due to its ability to accelerate training through better utilization of parallel processing power and reduction of stochastic variance. Another popular technique is importance sampling -- a strategy for prefer…
Develops minibatch stochastic proximal gradient for large-scale learning models.
problem Finding optimal predictors with complex regularizers in large-scale learning models.
method Minibatch variants of stochastic proximal gradient algorithm for composite objective functions.
result Minibatch size N after O(Nε1) iterations achieves ε−suboptimality in expected quadratic distance. Minibatch SGD outperforms Local SGD in heterogeneous distributed learning.
problem Optimizing a combined convex objective with stochastic gradient estimates from different machines.
method Analysis of Minibatch SGD and Local SGD in a heterogeneous distributed setting.
result Minibatch SGD dominates Local SGD in the heterogeneous distributed setting.
Local SGD outperforms minibatch SGD for quadratic objectives.
problem Theoretical foundations of local SGD are lacking.
method Proved local SGD strictly dominates minibatch SGD for quadratic objectives and accelerated local SGD is minimax optimal.
result Local SGD does not dominate minibatch SGD in general convex objectives.
Although stochastic gradient descent (SGD) is a driving force behind the recent success of deep learning, our understanding of its dynamics in a high-dimensional parameter space is limited. In recent years, some researchers have used the stochasticity of minibatch gradients, or the signal-to-noise ratio, to better char…
Scalable training of inference networks for Gaussian-process models.
problem Computational challenges in inference for large data in Gaussian process models.
method Stochastic inference networks with a functional mirror-descent algorithm for minibatch training.
result Empirical results show comparable and sometimes superior performance to existing sparse variational GP methods.
New minibatching strategy reduces stochastic gradient bias in optimisation.
problem Reducing bias in stochastic gradient descent.
method Symmetric Minibatching Strategy combined with momentum.
result Reduced stochastic gradient bias from O(h2) to O(h4). New algorithms accelerate model-based optimization for stochastic problems.
problem Optimizing model-based stochastic optimization problems efficiently.
method Proposed new model-based algorithms with acceleration and minibatch techniques.
result Non-asymptotic convergence guarantees with linear speedup in minibatch size.
Adaptive batch sizes improve local gradient methods in distributed training.
problem Communication bottlenecks in distributed deep learning.
method Adaptive batch size strategies for local gradient methods.
result Adaptive batch sizes reduce minibatch gradient variance and improve training efficiency.
New method reduces discrete flow transitions, improving perplexity estimation.
problem Stochasticity in discrete paths makes rectification strategies ineffective.
method Dynamic-optimal-transport-like minimization objective with minibatch strategies.
result 32 times reduction in transitions for same perplexity.
New algorithm samples Bayesian neural networks for improved calibration.
problem Improving calibration of Bayesian neural networks.
method Symmetric Minibatch Splitting-UBU (SMS-UBU) algorithm.
result SMS-UBU provides better calibration performance than standard methods.
New method shows stochastic momentum can converge quickly on optimization problems.
problem Improving convergence of stochastic optimization methods.
method Stochastic heavy ball momentum with minibatching.
result Stochastic heavy ball momentum retains fast linear rate on quadratic problems.
Study on the noise in SGD minibatches near local minima.
problem Understanding the noise in SGD minibatches near local minima.
method Detailed analysis of SGD noise in linear regression and derivation of a general formula for different types of minima.
result Provides insight into the stability of training neural networks and suggests large learning rates can help generalization.
A new method to decompose audio scenes using deep learning.
problem Understanding audio scenes with random microphone arrangements.
method Formulated a neural network for nonnegative tensor factorization.
result Learned sources' individual spectral dictionaries and activation patterns.
Paper examines the structure of stochastic gradients in deep learning.
problem Exploring the structure and heavy tails of stochastic gradients in deep learning.
method Conducted formal statistical tests on stochastic gradients and gradient noise.
result Stochastic gradients and gradient noise do not exhibit power-law heavy tails, but their covariance spectra do.
Paper proves minibatch SGD for GP inference converges and improves generalization.
problem Theoretical understanding and practical use of SGD for correlated samples in Gaussian process inference.
method Proves minibatch SGD converges to a critical point with rate O(1/K) for K iterations, under certain kernel conditions.
result Minibatch SGD for GP inference improves generalization and reduces computational burden.
SGD converges to global minimum for structured non-convex functions.
problem Optimizing non-convex functions using SGD with slow convergence rates.
method Convergence theorems for SGD on structured non-convex functions, including Quasar and PL conditions.
result SGD converges to global minimum for specific non-convex functions under certain conditions.
Noise enhancement improves generalization in training.
problem Improving generalization in training with controlled noise.
method Noise enhancement method to control SGD noise without changing learning rate or minibatch size.
result Noise enhancement improves generalization for real datasets.
New sampling technique improves KGC model performance.
problem Ignoring entity neighbors in minibatches affects KGC model training.
method Random-walk based minibatch sampling.
result Proposed method achieves state-of-the-art performance on DB100K.
We present novel minibatch stochastic optimization methods for empirical risk minimization problems, the methods efficiently leverage variance reduced first-order and sub-sampled higher-order information to accelerate the convergence speed. For quadratic objectives, we prove improved iteration complexity over state-of-…
Poor (even random) starting points for learning/training/optimization are common in machine learning. In many settings, the method of Robbins and Monro (online stochastic gradient descent) is known to be optimal for good starting points, but may not be optimal for poor starting points -- indeed, for poor starting point…
A new method for distributed optimization reduces communication rounds without minibatches.
problem Efficient training in distributed machine learning with different data distributions.
method A primal-dual method (GA-MSGD) applied to the Lagrangian of distributed optimization.
result Achieves linear convergence in communication rounds for strongly convex objectives.
Unified analysis of stochastic gradient methods for convex and smooth optimization.
problem Minimizing composite convex and smooth functions.
method Unified convergence analysis of various stochastic gradient methods.
result Unified convergence rates for a variety of methods including proximal SGD, variance reduced methods, quantization, and coordinate descent.
New bounds for M-SGD show its error distribution is nearly Gaussian.
problem Understanding the error distribution of M-SGD.
method Proved non-asymptotic bounds for M-SGD in Wasserstein distance.
result Error distribution of M-SGD is approximately Gaussian.
New technique trains deep neural networks without normalization or minibatch statistics.
problem Training deep neural networks at high learning rates without normalization.
method Channel-wise zero-mean initialization and gradient modification to maintain common mode rejection.
result Achieves higher accuracy compared to batch normalization and shows minibatches are unnecessary.
New method improves convergence of SPP for convex optimization problems.
problem Stochastic optimization and robustness to SGD.
method Minibatch Stochastic Proximal Point (M-SPP) method with stability analysis.
result M-SPP achieves faster convergence rates under smoothness and quadratic growth conditions.
SGLRW improves robustness of stochastic gradient MCMC methods.
problem Sensitivity to minibatch size and gradient noise in stochastic-gradient MCMC methods.
method Proposes Stochastic Gradient Lattice Random Walk (SGLRW) with lattice-based discretization.
result SGLRW remains stable in regimes where SGLD fails, including heavy-tailed gradient noise.
This work improves SGD minibatch sampling using determinantal point processes based on orthogonal polynomials.
problem Improving variance reduction in stochastic gradient descent (SGD) for large datasets.
method Orthogonal polynomial-based determinantal point processes for sampling minibatches in SGD.
result DPP minibatches lead to a smaller mean square approximation error than uniform minibatches.
SGD efficiently learns the XOR function with near-optimal sample complexity.
problem Learning the XOR function with a 2-layer neural network.
method Minibatch SGD on a 2-layer neural network with ReLU activations, focusing on signal-finding and signal-heavy phases.
result Achieves population error o(1) with dextpolylog(d) samples. We analyze stochastic algorithms for optimizing nonconvex, nonsmooth finite-sum problems, where the nonconvex part is smooth and the nonsmooth part is convex. Surprisingly, unlike the smooth case, our knowledge of this fundamental problem is very limited. For example, it is not known whether the proximal stochastic gra…
EigenGame improves eigendecomposition by offering unbiased updates for larger datasets.
problem Minibatch bias in EigenGame limits convergence and parallelism.
method Proposed unbiased stochastic update for EigenGame.
result Asymptotic equivalence to EigenGame, greater parallelism, and improved performance.
DReg boosts large-batch SGD's generalization and convergence.
problem Large-batch SGD struggles with generalization in deep learning.
method DReg replicates a layer to encourage parameter diversity.
result DReg improves generalization and convergence with large-batch SGD.
Paper proposes a method to estimate variance reduction in DNN training using importance sampling.
problem Challenges in assessing variance reduction during DNN training using importance sampling.
method Proposes a method for estimating variance reduction using minibatches sampled under importance sampling.
result Demonstrates consistent reduction in variance, improved training efficiency, and enhanced model accuracy.
Contrastive Divergence (CD) and Persistent Contrastive Divergence (PCD) are popular methods for training the weights of Restricted Boltzmann Machines. However, both methods use an approximate method for sampling from the model distribution. As a side effect, these approximations yield significantly different biases and…
New approach speeds up optimization with repeated gradient steps on same batch.
problem Performance bottlenecks in massive parallel pipelines with large batch sizes.
method Data echoing, taking repeated gradient steps on the same batch.
result Data echoing affords speedups on curvature-dominated part of convergence rate.
We present an approach towards convex optimization that relies on a novel scheme which converts online adaptive algorithms into offline methods. In the offline optimization setting, our derived methods are shown to obtain favourable adaptive guarantees which depend on the harmonic sum of the queried gradients. We furth…
Advances few-shot classification by treating it as supervised learning and proposing new training techniques.
problem Formulating the ability of humans to learn from limited data in machine learning.
method Formulated few-shot classification as a supervised learning problem and introduced multi-episode and cross-way training techniques.
result Proposed training strategies accelerate the training process without accuracy loss.
Stochastic Gradient Descent (SGD) is a central tool in machine learning. We prove that SGD converges to zero loss, even with a fixed (non-vanishing) learning rate - in the special case of homogeneous linear classifiers with smooth monotone loss functions, optimized on linearly separable data. Previous works assumed eit…
Improved stochastic Halpern iteration for fixed-point approximation in normed spaces.
problem Approximating fixed-points of nonexpansive and contractive operators in normed finite-dimensional spaces.
method Stochastic Halpern iteration with minibatch, analyzing oracle complexity.
result Improved oracle complexity for nonexpansive operators, with a lower bound of Ω(ε−3). Stochastic gradient method converges as fast as deterministic for overparametrized models.
problem Convergence rate of stochastic gradient methods in overparametrized models.
method Proposes a regularity condition enabling fast convergence of SGD.
result Stochastic gradient method achieves the same convergence rate as deterministic gradient method.
New convergence bounds for shuffling-based SGD methods in distributed learning.
problem Analyzing the performance of shuffling-based variants of SGD in distributed learning.
method Study of minibatch and local Random Reshuffling methods, proving convergence bounds and lower bounds.
result Shuffling-based variants converge faster than with-replacement sampling methods, and the bounds are tight.
Noise in SGD helps deep nets generalize better, even with smaller batch sizes.
problem The generalization benefit of using noise in SGD over large batch sizes.
method Carefully designed experiments and rigorous hyperparameter sweeps on various models.
result Small or moderately large batch sizes outperform very large batches on test sets.
A scalable GPLVM model using stochastic variational inference.
problem Scalable inference for Gaussian process latent variable models.
method Doubly stochastic formulation of Bayesian GPLVM with minibatch training.
result High-fidelity reconstructions in the presence of missing data.
Stochastic gradient Markov Chain Monte Carlo (SG-MCMC) has been developed as a flexible family of scalable Bayesian sampling algorithms. However, there has been little theoretical analysis of the impact of minibatch size to the algorithm's convergence rate. In this paper, we prove that under a limited computational bud…