Polynomial-time algorithm for list-decodable linear regression with batches.
problem Efficiently decoding linear regression with a fraction of adversarial data.
method Polynomial time algorithm using batches of i.i.d. samples.
result Returns a list of size O(1/α^2) with one item close to true parameter.
New algorithm for batch list-decodable linear regression with stronger guarantees.
problem Efficiently list-decoding linear regression with a fraction of corrupted batches.
method Uses higher-order moments and Sum-of-Squares (SoS) certification to achieve better guarantees.
result Achieves substantially smaller minimum batch size and final error, with optimal list size.
Boosting improves accuracy with fewer calls to weak learners for certain concept classes.
problem Improving accuracy of learning algorithms with limited weak learner calls.
method Combines boosting and list-decodable codes to achieve better performance for specific concept classes.
result A new boosting algorithm that achieves strong learning with fewer calls to weak learners and additional samples.
New method for estimating sparse means in noisy data.
problem Estimating the mean of a sparse distribution in the presence of outliers.
method Difference-of-Pairs Filtering technique for list-decodable sparse mean estimation.
result First sample and computationally efficient algorithm for list-decodable sparse mean estimation.
Paper tackles high-accuracy list-decodable learning for mean estimation.
problem Estimate the mean of a distribution with a small fraction coming from a nice distribution.
method Developed a novel proof of identifiability and a new algorithmic approach.
result Achieved high-accuracy guarantees for list-decodable mean estimation.
New algorithm for robust mean estimation in high dimensions with nearly-PCA time complexity.
problem Robust mean estimation in high-dimensional data with a minority of contaminated data.
method List-decodable mean estimation algorithm using a novel soft downweighting method, SIFT, and a Ky Fan matrix multiplicative weights procedure.
result Optimal sample complexity and error rate for list-decodable mean estimation with runtime nearly matching PCA.
Statistical Query lower bound shows difficulty in list-decodable linear regression.
problem List-decodable linear regression with adversarial corruption.
method Statistical Query (SQ) lower bound analysis.
result Lower bound of dpoly(1/α) for list-decodable linear regression. Polynomial-time algorithm for estimating covariance in corrupted Gaussian data.
problem Estimating covariance in data with up to 1-α fraction of adversarial corruptions.
method Uses low-degree sum-of-squares certificates for anti-concentration and hypercontractivity.
result Outputs a list of candidate parameters with high probability containing a nearly correct covariance.
New algorithm recovers mixture means even with many outliers.
problem Estimating mixture means when outliers overwhelm small groups.
method Proposes an algorithm for robust mixture learning with minimal list-size overhead.
result Order-optimal error guarantees for each mixture mean.
Algorithm estimates covariance from noisy data efficiently.
problem Estimating covariance from a noisy set of points.
method Spectral techniques for list-decodable covariance estimation.
result Efficient algorithm with poly(1/α) sample and time complexity.
We give the first polynomial-time algorithm for robust regression in the list-decodable setting where an adversary can corrupt a greater than 1/2 fraction of examples. For any α<1, our algorithm takes as input a sample {(xi,yi)}i≤n of n linear equations where αn of the equations satisfy $y_i = \l…
A simple model explains phase transition in large language models.
problem Understanding the emergence of abilities in large language models.
method Modeling LLM as a sequence-to-sequence random function and using a list decoder.
result A critical threshold exists where the expected number of erroneous sequences grows exponentially.
New method estimates mean from noisy data with few outliers.
problem Estimate mean from noisy data with few outliers.
method Iterative multi-filtering approach.
result Achieves near-optimal error with practical efficiency.
New method certifies anti-concentration for various non-Gaussian distributions.
problem Efficiently certifying anti-concentration for non-Gaussian distributions.
method Sum-of-Squares relaxation of integer program for anti-concentration.
result Quasi-polynomial time certificates for non-Gaussian distributions.
New algorithm clusters data with almost-linear time, robust to corruption.
problem Clustering mixtures of distributions with adversarial corruption.
method List-decodable mean estimation, robust mean estimation, iterative multi-filtering.
result First almost-linear time algorithms for clustering mixtures of k well-behaved distributions.
Improved subspace recovery algorithm with dimension-independent error and polynomial time.
problem Efficiently recover a covariance matrix from a mix of inliers and adversarial outliers.
method List-decodable subspace recovery algorithm with faster fixed-polynomial time and less restrictive distributional assumptions.
result Achieved dimension-independent error guarantee of O(1/α) with poly(1/α d^O(1)) time complexity.
Paper tackles MLR prediction error without assuming realizable models.
problem Prediction error in mixture of linear regressions without realizable assumptions.
method Developed algorithms for list-decoding MLR predictions and minimized empirical risk.
result Alternating minimization algorithm finds best fit lines in non-realizable settings.
New research shows many batch selection methods for training work just as well as full batch training.
problem Finding optimal batch selection methods for training.
method Analysis of mini-batch Gradient Descent (GD) and Stochastic GD (SGD) with various batch selection rules.
result All mini-batch schedules, including deterministic ones, generalize optimally for smooth Lipschitz-convex/nonconvex/strongly-convex loss functions.
Determining the appropriate batch size for mini-batch gradient descent is always time consuming as it often relies on grid search. This paper considers a resizable mini-batch gradient descent (RMGD) algorithm based on a multi-armed bandit for achieving best performance in grid search by selecting an appropriate batch s…
We analyze how batch learning impacts bandit problems.
problem Impact of batch learning in stochastic bandits.
method Policy-agnostic regret analysis, upper and lower bounds demonstration.
result The impact of batch learning can be measured in terms of online behavior.
BatchGFN uses generative flow networks for efficient batch active learning.
problem Efficiently selecting informative batches for active learning.
method Generative flow networks to sample batches proportional to a batch reward.
result Constructs highly informative batches with a single forward pass per point.
Adaptive batch sizes improve active learning efficiency and flexibility.
problem Fixed batch sizes in active learning are inefficient due to dynamic cost-speed trade-offs.
method Probabilistic Numerics framework that adaptively changes batch sizes based on integration error and precision objectives.
result Significant enhancement in learning efficiency and flexibility across various applications.
CBN improves batch normalization for small mini-batch sizes.
problem Reduced effectiveness of Batch Normalization in small mini-batch sizes.
method CBN uses statistics from multiple recent iterations, compensating for network weight changes via Taylor polynomials.
result CBN outperforms original batch normalization and direct iteration statistics in object detection and image classification.
Opt-BBAI identifies the best arm with minimal batches and pulls, optimizing both sample and batch complexity.
problem Batched best arm identification (BBAI) problem, aiming to minimize policy switches and resource usage.
method Proposed Opt-BBAI algorithm, achieving near-optimal sample and batch complexity in non-asymptotic settings.
result First algorithm to achieve near-optimal sample and batch complexity in non-asymptotic settings.
Batch Normalization (BN) uses mini-batch statistics to normalize the activations during training, introducing dependence between mini-batch elements. This dependency can hurt the performance if the mini-batch size is too small, or if the elements are correlated. Several alternatives, such as Batch Renormalization and G…
A new method designs batches for Bayesian optimization more efficiently.
problem Efficiently designing batches for Bayesian optimization to reduce total time.
method Minimal Terminal Variance (MTV) acquisition function, optimizing I-optimality criterion.
result MTV designs batches more efficiently than other methods, as shown by numerical experiments.
Training deep neural networks with Stochastic Gradient Descent, or its variants, requires careful choice of both learning rate and batch size. While smaller batch sizes generally converge in fewer training epochs, larger batch sizes offer more parallelism and hence better computational efficiency. We have developed a n…
Riemannian stochastic gradient descent converges faster with increasing batch size.
problem Improving convergence rate of Riemannian stochastic gradient descent.
method Theoretical analysis and numerical investigation of increasing batch size effects.
result Riemannian stochastic gradient descent converges faster with increasing batch size.
New findings show mini-batch SGD operates in a 'Edge of Stochastic Stability' regime.
problem Understanding the stability and convergence of mini-batch SGD.
method Analyzing the mini-batch Hessian and its directional curvature.
result Mini-batch SGD operates in a different stability regime (Edge of Stochastic Stability) compared to full-batch GD.
Batch normalisation doesn't affect variational inference but fails for larger batch sizes.
problem Failure of Monte Carlo Batch Normalisation (MCBN) for capturing epistemic uncertainty in larger batch sizes.
method Investigated MCBN as an approximate inference technique for Bayesian neural networks, showing its limitations and providing insights for improvement.
result For larger batch sizes, MCBN fails to capture epistemic uncertainty, requiring the batch size to be a variational parameter.
A new scaling law predicts optimal batch size for training models.
problem Finding the optimal batch size for training models efficiently.
method Proposed a three-term scaling law that considers model size, training data, training steps, and batch size.
result The three-term law accurately recovers the optimal batch size and can be robustly fit with fewer training runs.
E4 algorithm optimizes batched linear bandits with minimal regret and batches.
problem Optimizing batched linear bandits for minimal regret.
method Explore-Estimate-Eliminate-Exploit framework with optimal exploration rate.
result Achieves minimax and asymptotic optimality in regret and batch complexity.
ABS dynamically adjusts batch size based on policy stability, improving RL performance.
problem Diminishing returns with large batch sizes in RL due to non-stationary data.
method Adaptive Batch Scaling (ABS) with Behavioral Divergence metric.
result Larger batch sizes can improve RL performance, contrary to conventional wisdom.
SNGM improves large-batch training accuracy.
problem Improving generalization in large-batch training.
method Stochastic Normalized Gradient Descent with Momentum.
result SNGM achieves better test accuracy than MSGD and other large-batch methods.
Batched Neural Bandits reduces policy updates in sequential decision-making.
problem Sequential decision-making with batched policy changes.
method BatchNeuralUCB algorithm combining neural networks and optimism.
result Achieves similar regret as fully sequential version with fewer policy updates.
Large batch sizes reduce gradient variance in DP-SGD, improving privacy.
problem Understanding why large batch sizes work in DP-SGD.
method Decomposed total gradient variance into subsampling and noise-induced variances, proving batch size independence in the limit.
result Large batch sizes reduce effective total gradient variance, improving privacy in DP-SGD.
Large-batch SGD is important for scaling training of deep neural networks. However, without fine-tuning hyperparameter schedules, the generalization of the model may be hampered. We propose to use batch augmentation: replicating instances of samples within the same batch with different data augmentations. Batch augment…
The paper refines and extends batched kernelized bandits, improving regret bounds and introducing a robust setting.
problem Optimizing black-box functions with noisy batches in Reproducing Kernel Hilbert Space.
method Refined and extended existing regret bounds, including adaptive batch sizes and robust optimization.
result Improved regret bounds for batched kernelized bandits, showing optimal number of batches and adaptive batch sizes.
Batch Thompson Sampling reduces exploration-exploitation trade-off in online decision making.
problem Balancing exploration and exploitation in online decision making.
method Introducing a batch Thompson Sampling framework for stochastic multi-arm bandit and linear contextual bandit problems.
result Achieves asymptotic regret bound with O(logT) batch queries, significantly reducing interactions. A new method SEBS optimizes SGD batch size for better performance.
problem Optimizing batch size for SGD to balance training speed and generalization.
method SEBS method uses a multi-stage geometric batch size enlargement scheme.
result SEBS reduces parameter updates without increasing generalization error.
Improved batched SH algorithm maintains original performance.
problem Maintaining performance in batched multi-armed bandits.
method Simple batch version of Sequential Halving algorithm.
result Batching does not degrade SH algorithm's performance.
Paper tackles robust batched bandits for heavy-tailed rewards.
problem Clinical trials and other applications with heavy-tailed rewards.
method Proposes robust batched bandit algorithms for heavy-tailed rewards in finite-arm and Lipschitz-continuous settings.
result Heavier-tailed rewards require fewer batches for near-optimal regret in the instance-independent regime and Lipschitz setting.
The study analyzes batched methods for early stopping in stochastic multi-armed bandits.
problem Early stopping in stochastic multi-armed bandits with fixed confidence.
method Instance-dependent lower bounds and a general batched algorithm with upper bounds.
result Upper and lower bounds on the number of batches and sample complexity.
Distributed implementations of mini-batch stochastic gradient descent (SGD) suffer from communication overheads, attributed to the high frequency of gradient updates inherent in small-batch training. Training with large batches can reduce these overheads; however, large batches can affect the convergence properties and…
This work studies the implicit bias of mini-batch SGD in classification.
problem Understanding the implicit bias of mini-batch SGD in multi-class classification.
method Characterizes how batch size, momentum, and variance reduction affect convergence and max-margin behavior under different norms.
result Momentum enables small-batch convergence to an approximate max-margin solution, while variance reduction recovers the exact full-batch bias.
A new method normalizes activations to match batch normalization without batch dependence.
problem Performance degradation with batch-independent normalization techniques.
method Proxy-Normalizing Activations
result Proxy-Normalization technique emulates batch normalization's behavior and performance.
Algorithm achieves comparable performance to fully dynamic data with only a few batches.
problem High-dimensional multi-armed contextual bandits with batched feedback.
method Provable sample-efficient algorithm using batch allocation method.
result Achieves regret bounds comparable to fully sequential setting with only L = O(log T) batches.
Analysis of SGD+M convergence rates in high dimensions with batch size considerations.
problem Understanding convergence rates of SGD+M in high-dimensional settings.
method Analyzing the dynamics of SGD+M on least squares problems with large batch sizes and dimensions.
result Identifies the implicit conditioning ratio (ICR) that regulates SGD+M's acceleration and convergence rates.