Reduces deep learning training data for faster testing.
problem Resource-intensive deep learning training with full data sets.
method Evaluated different training set reduction methods.
result Training set reduction is useful in resource-constrained environments.
MARS optimizes large model training by reducing variance, outperforming AdamW.
problem Training large models efficiently and scalably.
method Unified optimization framework MARS combining preconditioned gradient updates and variance reduction.
result MARS outperforms AdamW in training GPT-2 models.
We present local discriminative Gaussian (LDG) dimensionality reduction, a supervised dimensionality reduction technique for classification. The LDG objective function is an approximation to the leave-one-out training error of a local quadratic discriminant analysis classifier, and thus acts locally to each training po…
Active sampling selects few points for accurate model reduction of high-fidelity systems.
problem Efficiently identify dominant subspaces for model reduction of large training sets.
method Proposes an active sampling strategy to select a few points from the training set to estimate dominant subspaces accurately.
result Active sampling can provide 17x speed-up without sacrificing accuracy.
VR-ConfTr reduces noise in CP training, leading to more stable and efficient model performance.
problem Improving the efficiency and stability of conformal prediction during model training.
method Variance-reduced conformal training (VR-ConfTr) that incorporates variance reduction in gradient estimation.
result VR-ConfTr achieves faster convergence and smaller prediction sets compared to existing methods.
K-Nearest neighbor classifier (k-NNC) is simple to use and has little design time like finding k values in k-nearest neighbor classifier, hence these are suitable to work with dynamically varying data-sets. There exists some fundamental improvements over the basic k-NNC, like weighted k-nearest neighbors classifier (wh…
Enhances supervised visualization for unseen data using autoencoders and random forest.
problem Lack of generalization to unseen test sets in supervised dimensionality reduction.
method Combines autoencoder and random forest proximities for out-of-sample extension.
result 40% reduction in training time with 10% of training data, achieving consistent quality.
The diverse world of machine learning applications has given rise to a plethora of algorithms and optimization methods, finely tuned to the specific regression or classification task at hand. We reduce the complexity of algorithm design for machine learning by reductions: we develop reductions that take a method develo…
A neural network approach for feature selection using mutual information.
problem Feature ranking and selection leading to sub-optimal solutions for class separability.
method Stochastic mutual information gradient estimation for dimensionality reduction.
result The network projects features onto an output space maximizing mutual information with class labels.
How can we train a statistical mixture model on a massive data set? In this work we show how to construct coresets for mixtures of Gaussians. A coreset is a weighted subset of the data, which guarantees that models fitting the coreset also provide a good fit for the original data set. We show that, perhaps surprisingly…
Graphical heuristic reduces and partitions large datasets for faster supervised training.
problem Training large datasets for classification tasks.
method Clustering and information graph construction for dataset reduction and partitioning.
result Significant speed-up in training run-time without compromising prediction accuracy.
Communication on heterogeneous edge networks is a fundamental bottleneck in Federated Learning (FL), restricting both model capacity and user participation. To address this issue, we introduce two novel strategies to reduce communication costs: (1) the use of lossy compression on the global model sent server-to-client;…
This paper studies how key tensor properties are inherited in subtensors of tensor train decompositions.
problem Theoretical development of property inheritance for subtensors in tensor train decompositions.
method Theoretical analysis of incoherence and condition number preservation, and tensor train rank preservation through fiber-wise sampling.
result Key tensor properties (incoherence and condition number) can be well preserved to subtensors formed via fiber-wise sampling.
Sequential prediction problems such as imitation learning, where future observations depend on previous predictions (actions), violate the common i.i.d. assumptions made in statistical learning. This leads to poor performance in theory and often in practice. Some recent approaches provide stronger guarantees in this se…
RMDA trains structured neural networks with regularization and variance reduction.
problem Training structured neural networks with desired properties.
method RMDA algorithm for structured NNs with regularization and variance reduction.
result RMDA achieves desired structures identical to regularizer's at stationary points.
A new distributed SGD algorithm reduces communication by infrequent global reduction.
problem Reducing communication in large-scale machine learning training.
method Hierarchical averaging stochastic gradient descent (Hier-AVG) with infrequent global reduction.
result Hier-AVG achieves comparable training speed with better test accuracy.
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.
A new PMA method combines PCA and PLS for better classification.
problem Improving classification performance in data analysis.
method Combines PCA and PLS through multiple sub-PLS models and PCA on the joint coefficient matrix.
result The proposed PMA method achieves better classification performance and stability.
In the covariate shift learning scenario, the training and test covariate distributions differ, so that a predictor's average loss over the training and test distributions also differ. In this work, we explore the potential of extreme dimension reduction, i.e. to very low dimensions, in improving the performance of imp…
RSAC improves lightweight continuous learning efficiency.
problem Excessive training time and memory usage in continuous learning.
method Regularized subspace approximation classifier with feature reduction and regularization.
result RSAC achieves more efficient continuous learning than prior methods.
We present a theoretical analysis and empirical evaluations of a novel set of techniques for computational cost reduction of classifiers that are based on learned transform and soft-threshold. By modifying optimization procedures for dictionary and classifier training, as well as the resulting dictionary entries, our t…
The paper studies a rebalanced dataset for imbalanced classification using Centered Random Forests.
problem Imbalanced classification where one class is underrepresented.
method Theoretical analysis of Centered Random Forests (CRF) with rebalanced datasets and debiasing techniques.
result Theoretical Central Limit Theorem (CLT) for the infinite CRF and debiased estimator IS-ICRF.
AI model identifies chemical agents in MCI with high accuracy.
problem Identifying chemical agents in mass casualty incidents.
method Reverse engineered signs/symptoms, trained using ANN, BDT, and WISER.
result WISER outperformed ANN and BDT in identifying chemical agents.
A novel deep learning method for chemometric data improves performance over transfer learning.
problem Training deep neural networks from chemometric data with varying input sizes.
method Weight sharing in deep convolutional neural networks trained on multiple data sets of different sizes.
result Superior performance compared to transfer learning, especially when training on medium and small data sets.
Paper improves tree probability estimation using stochastic optimization and variance reduction.
problem Improving tree probability estimation in phylogenetic inference.
method Introduces computationally efficient methods for training SBNs and variance reduction for optimization.
result Methods outperform previous baseline methods in tree topology probability estimation and Bayesian phylogenetic inference.
RMFGP combines multi-fidelity models for efficient uncertainty quantification.
problem Efficiently infer quantities of interest with limited high-fidelity data.
method Rotated multi-fidelity Gaussian process with dimension reduction and Bayesian active learning.
result RMFGP model improves accuracy and efficiency in high-dimensional problems.
New method for multi-modal depth prediction challenges.
problem Challenges in applying unimodal coreset selection to multi-modal data.
method Adapted state-of-the-art coreset selection technique for multimodal data.
result Challenges in extending unimodal algorithms to multi-modal scenarios.
TTPUDR uses tensor-train decomposition for high-dimensional data analysis.
problem High-dimensional data analysis challenges.
method Tensor-train decomposition and manifold optimization.
result TTPUDR significantly outperforms past methods and state-of-the-art methods.
Locally private Naive Bayes works for personal data.
problem Training Naive Bayes on personal data with privacy concerns.
method Local differential privacy, dimensionality reduction, and perturbation techniques.
result Naive Bayes accuracy maintained under local differential privacy.
KD technique improves QDNN performance with reduced hyper-parameters.
problem Restoring performance loss in QDNNs due to quantization.
method Applied KD with reduced hyper-parameters, including a new coefficient reduction technique.
result Achieved 92.7% test accuracy on CIFAR-10 and 67.0% on CIFAR-100 with 2-bit weights.
New optimizer MARS-M combines variance reduction with Muon for faster LLM training.
problem Training large-scale neural networks efficiently.
method Integrates MARS variance reduction with Muon optimizer.
result MARS-M converges to a first-order stationary point at a rate of ildeO(T−1/3). LIIT uses large deviations to train neural networks faster.
problem Training neural networks requires large amounts of data, leading to inefficiencies.
method LIIT uses large deviations principle to generate and update training samples.
result LIIT achieves similar classification performance with smaller, more representative samples.
PosCal training improves classification models by calibrating posterior probabilities.
problem Poorly calibrated posterior probabilities in classification models.
method End-to-end training procedure that directly optimizes the objective while minimizing the difference between predicted and empirical posterior probabilities.
result PosCal training achieves about 2.5% task performance gain and 16.1% calibration error reduction.
WeldNet reduces complex dynamics to simpler, manageable segments.
problem Complex, high-dimensional time-dependent datasets from physical processes are costly to simulate.
method Windowed Encoders for Learning Dynamics, splitting time domain into windows for nonlinear dimension reduction and propagator training.
result WeldNet captures nonlinear latent structures and dynamics, outperforming existing methods.
Deep networks become equivalent to linear models in large data regimes.
problem Understanding the behavior of deep neural networks in large data regimes.
method Information-theoretic analysis of fully-trained neural networks in proportional scaling regime.
result Proves deep Gaussian equivalence principle, showing deep networks can be simplified to linear models.
Mime algorithm improves federated learning by adapting centralized methods.
problem Mitigating client drift in federated learning.
method Combines control variates and server-level statistics to adapt centralized algorithms to federated learning.
result Mime outperforms any centralized method in federated learning.
Abstract: Generalized reduction methods for symmetries in graded geometry.
problem Generalized reduction of symmetries in graded geometry.
method Graded symplectic reduction for Courant, Dirac, and generalized complex structures.
result Systematic recovery of reduction schemes for exact cases.
New method finds efficient low-rank neural networks during training.
problem High memory and computational demands of neural networks.
method Restricts weight matrices to a low-rank manifold and updates low-rank factors.
result Significantly reduced time and memory resources required for training and evaluation.
Framework transfers complementary operating conditions to train anomaly detectors.
problem Training anomaly detectors on changing operating conditions requires comprehensive data, which is hard to obtain.
method Proposes unsupervised transfer learning to align and combine data from different units.
result Demonstrates improved anomaly detection in changing operating conditions.
Reduces function approximation dimensions from high to low with sparse data.
problem Function approximation from sparse data.
method Nonlinear Level Set Learning (NLL) with geometric information.
result Reduces input dimension to theoretical lower bound with minor accuracy loss.
New method finds Nash equilibria faster in multiplayer games.
problem Finding Nash equilibria in multi-player games, especially in noisy environments.
method Extra-gradient with player sampling and variance reduction.
result Proves better convergence rate than full extra-gradient for noisy games.
Dimensionality reduction is a topic of recent interest. In this paper, we present the classification constrained dimensionality reduction (CCDR) algorithm to account for label information. The algorithm can account for multiple classes as well as the semi-supervised setting. We present an out-of-sample expressions for …
New method ensures consistent inference across different tensor parallel sizes for large language models.
problem Non-deterministic inference in large language models due to inconsistent reduction orders across GPUs.
method Tree-Based Invariant Kernels (TBIK) that align intra- and inter-GPU reduction orders through a unified hierarchical binary tree structure.
result Bit-wise identical results across different tensor parallel sizes for RL training.
Parametric t-SNE improves generalization for streaming data.
problem Training neural networks for t-SNE objective function fails due to gradient exploding.
method Applied gradient clipping to solve gradient exploding problem.
result Parametric t-SNE achieves quality compatible with non-parametric t-SNE while generalizing to new data.
New algorithms improve SGD convergence and reduce variance for over-parameterized models.
problem Slower convergence in non-interpolation settings for SGD variants.
method Proposed AdaSPS and AdaSLS with variance reduction for robust convergence.
result Achieves faster convergence rates and robustness in non-interpolation settings.
TTRP method preserves distances in high-dimensional data with reduced storage and speed.
problem Preserving distances in high-dimensional datasets efficiently and accurately.
method Tensor train random projection (TTRP) using TT-ranks of one.
result TTRP is an expected isometric projection with bounded variance.
Meta-CVs leverage task similarity to reduce variance with limited data.
problem Reducing variance in Monte Carlo estimators with few samples.
method Meta-learning control variates for related tasks.
result Meta-CVs lead to significant variance reduction in settings with limited data.
Efficient deep learning on exascale supercomputers solves materials imaging inverse problems.
problem Solving scientific inverse problems in materials imaging using deep learning.
method Novel communication strategies in synchronous distributed deep learning, including decentralized gradient reduction and computational graph-aware grouping.
result Achieved near-linear scaling of distributed training up to 27,600 GPUs on Summit, reaching 2.15(4) EFLOPS16.