Linear algebra approach for parallel deep learning models.
problem Training large DNNs in distributed environments.
method Linear algebraic approach to model parallelism.
result Manual development of backward operators for gradient-based training.
On-device research index
arXiv research
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
168,742 papers · 148 categories
Trend · papers per month
48 results for “parallel distributions”
Method predicts how probability distributions evolve over time.
problem Predicting how systems described by probability distributions evolve under different conditions.
method Wasserstein Parallel Transport
result Wasserstein Parallel Transport provides counterfactual comparisons of distributional dynamics.
New method accelerates Parallel Tempering using neural samplers.
problem Challenges in sampling from high-dimensional, multimodal distributions.
method Leverages neural samplers to reduce overlap between distributions.
result Improves sample quality and reduces computational cost.
TensorOpt finds optimal parallelization strategies for DNN training.
problem Finding efficient parallelization strategies for DNN training.
method TensorOpt uses an algorithm (FT) to search for an optimal set of parallelization strategies considering multiple objectives.
result TensorOpt provides accurate runtime cost estimation and adapts to resource availability.
New adaptive temperature selection improves parallel tempering efficiency.
problem Enhancing mixing in multi-modal distributions using parallel tempering.
method Adaptive temperature selection using policy gradient approach.
result Lower integrated autocorrelation times achieved compared to traditional methods.
Proposes a method to improve SLMC for multimodal distributions.
problem Difficulty of applying SLMC to multimodal distributions.
method Parallel adaptive annealing with VAE-SLMC.
result Can proficiently obtain accurate samples from multimodal distributions.
Parallel sampling for smooth distributions with fast convergence.
problem Efficiently sampling from distributions with smooth densities.
method Parallelization of Langevin algorithms under log-Sobolev inequalities.
result Samples close to target distribution with low KL divergence or TV distance.
Embarrassingly (communication-free) parallel Markov chain Monte Carlo (MCMC) methods are commonly used in learning graphical models. However, MCMC cannot be directly applied in learning topic models because of the quasi-ergodicity problem caused by multimodal distribution of topics. In this paper, we develop an embarra…
Study examines parallel computing strategies for faster imputation of missing data.
problem Time-consuming iterative imputation methods for large datasets.
method Variable-wise and model-wise distributed parallel computing strategies in missForest.
result Variable-wise distributed strategy introduces additional biases in imputation results.
An infinite parallel tempering bouncy particle sampler improves sampling efficiency for multimodal distributions.
problem Sampling from complex posterior distributions with high accuracy and efficiency.
method Introduced an infinite parallel tempering bouncy particle sampler (BPS-PT) to accelerate convergence.
result Demonstrated improved sampling efficiency for multimodal distributions through numerical simulations.
Enhances gradient-based discrete samplers with parallel tempering for multimodal distributions.
problem Local minima in high-dimensional, multimodal discrete distributions.
method Combines parallel tempering with discrete Langevin proposal, using Metropolis criterion for swaps.
result Significantly faster mixing and better sampling from complex distributions.
This paper studies parallelization schemes for stochastic Vector Quantization algorithms in order to obtain time speed-ups using distributed resources. We show that the most intuitive parallelization scheme does not lead to better performances than the sequential algorithm. Another distributed scheme is therefore intro…
New method uses higher-order Langevin dynamics for efficient parallel sampling.
problem Efficient parallel sampling from high-dimensional log-concave distributions.
method Combines higher-order Langevin dynamics with blockwise Lagrange polynomial interpolation.
result Reduces the number of parallel points required for a target accuracy.
Clapping reduces memory usage in distributed optimization by reusing data samples.
problem Significant communication overhead and impractical memory overhead in pipeline-parallel distributed optimization.
method Lazy sampling strategy to reuse data samples across steps, supporting convergence without unbiased gradient assumptions.
result Clapping achieves convergence in few-epoch or online training regimes without sample-size memory overhead.
Parallel and Distributed Thompson Sampling for Large-scale Accelerated Exploration of Chemical Spacestat.ML
Chemical space is so large that brute force searches for new interesting molecules are infeasible. High-throughput virtual screening via computer cluster simulations can speed up the discovery process by collecting very large amounts of data in parallel, e.g., up to hundreds or thousands of parallel measurements. Bayes…
DistShap parallelizes GNN explanation for large graphs.
problem Computational expense in attributing GNN predictions to specific edges or features.
method Distributed Shapley values across multiple GPUs for scalable GNN explanations.
result DistShap outperforms existing methods and scales to models with millions of features.
Gaussian processes (GP) are Bayesian non-parametric models that are widely used for probabilistic regression. Unfortunately, it cannot scale well with large data nor perform real-time predictions due to its cubic time cost in the data size. This paper presents two parallel GP regression methods that exploit low-rank co…
Gaussian processes (GP) are Bayesian non-parametric models that are widely used for probabilistic regression. Unfortunately, it cannot scale well with large data nor perform real-time predictions due to its cubic time cost in the data size. This paper presents two parallel GP regression methods that exploit low-rank co…
ESS improves MCMC efficiency for correlated & multimodal distributions.
problem Slice Sampling's sensitivity to initial length scale and difficulty with correlated distributions.
method Adaptive tuning and parallel walkers for efficient sampling.
result ESS improves efficiency by more than an order of magnitude on correlated distributions.
We consider parallel asynchronous Markov Chain Monte Carlo (MCMC) sampling for problems where we can leverage (stochastic) gradients to define continuous dynamics which explore the target distribution. We outline a solution strategy for this setting based on stochastic gradient Hamiltonian Monte Carlo sampling (SGHMC) …
EP-GFlowNets parallelize GFlowNet training for large-scale Bayesian inference.
problem Prohibitive repeated evaluations of unnormalized distributions for large-scale posterior sampling.
method Divide-and-conquer approach with server learning from local models.
result EP-GFlowNets enable efficient parallel and federated Bayesian inference.
Bayesian matrix factorization (BMF) is a powerful tool for producing low-rank representations of matrices and for predicting missing values and providing confidence intervals. Scaling up the posterior inference for massive-scale matrices is challenging and requires distributing both data and computation over many worke…
HPSGD speeds up DNN training by paralleling data sync with local training.
problem Low cluster utilization in distributed deep neural network training.
method Hierarchical Parallel SGD (HPSGD) with improved model updating for stale gradients.
result Significantly boosts distributed DNN training and reduces stale gradients.
To scale non-parametric extensions of probabilistic topic models such as Latent Dirichlet allocation to larger data sets, practitioners rely increasingly on parallel and distributed systems. In this work, we study data-parallel training for the hierarchical Dirichlet process (HDP) topic model. Based upon a representati…
Study on null-projectability of Levi-Civita connections in neutral metrics.
problem Characterizing projectability of Levi-Civita connections along null parallel distributions.
method Analyzing projectability of torsion-free connections along foliations on manifolds, focusing on neutral metric signatures and mid-dimensional distributions.
result Extension of Patterson and Walker's Riemann extension metrics to null parallel distributions of any dimension.
MindFlayer SGD improves parallel SGD for heterogeneous, random compute times.
problem Minimizing nonconvex functions with heterogeneous, random compute times.
method MindFlayer SGD, designed for stochastic and heterogeneous delays.
result MindFlayer SGD outperforms existing methods in environments with heavy-tailed noise.
Improves parallel deep model performance by restructuring and pruning.
problem Latency in parallel deep model execution due to interdependency among sub-models.
method Layer-wise model restructuring and pruning, using optimization and Munkres assignment algorithm.
result Significantly improves efficiency of distributed inference in terms of communication and computational complexity.
Batch-splitting (data-parallelism) is the dominant distributed Deep Neural Network (DNN) training strategy, due to its universal applicability and its amenability to Single-Program-Multiple-Data (SPMD) programming. However, batch-splitting suffers from problems including the inability to train very large models (due to…
On the local structure of Lorentzian Einstein manifolds with parallel distribution of null linesmath.DG
We study transformations of coordinates on a Lorentzian Einstein manifold with a parallel distribution of null lines and show that the general Walker coordinates can be simplified. In these coordinates, the full Lorentzian Einstein equation is reduced to equations on a family of Einstein Riemannian metrics.
When training large machine learning models with many variables or parameters, a single machine is often inadequate since the model may be too large to fit in memory, while training can take a long time even with stochastic updates. A natural recourse is to turn to distributed cluster computing, in order to harness add…
PALMS reconstructs large-scale networks efficiently with parallel computing.
problem Reconstructing large-scale latent networks from observed dynamics is computationally challenging.
method PALMS (Parallel Adaptive Lasso with Multi-directional Signals) framework for distributed network reconstruction.
result PALMS substantially reduces computational complexity and storage requirements.
Adaptive batch size schedules improve language model training efficiency and generalization.
problem Dilemma of choosing batch sizes in large-scale model training.
method General-purpose adaptive batch size schedules compatible with data and model parallelism.
result Adaptive batch size schedules outperform constant batch sizes and heuristic warmup schedules.
Communication-Efficient, 2D Parallel Stochastic Gradient Descent for Distributed-Memory Optimizationcs.DC
HybridSGD improves SGD performance by balancing computation and communication.
problem Limited scalability and performance of SGD due to communication costs.
method 2D parallel SGD method (HybridSGD) that trades off between 1D -step SGD and 1D Federated SGD (FedAvg).
result HybridSGD achieves better convergence than FedAvg at similar processor scales and up to 121x speedup over FedAvg.
NeLLoC improves image compression with parallel decoding.
problem Image compression with OOD generalization.
method Local autoregressive model with parallel decoding.
result Significant gains in compression runtime.
A new gradient quantization scheme improves communication efficiency in distributed training.
problem Efficiently compressing gradients for parallel training of large models.
method Proposes a new gradient quantization scheme with theoretical guarantees and empirical performance.
result The new scheme matches and exceeds the performance of existing methods.
We propose a new integrated method of exploiting model, batch and domain parallelism for the training of deep neural networks (DNNs) on large distributed-memory computers using minibatch stochastic gradient descent (SGD). Our goal is to find an efficient parallelization strategy for a fixed batch size using process…
Study explores geometric structure and prior for beta-logistic distribution.
problem Understanding the geometric structure and prior distributions of the beta-logistic distribution.
method Exploring dual geometric structure and uncovering -parallel prior.
result The beta-logistic distribution admits an -parallel prior for any real number .
A new sampler for complex discrete distributions efficiently updates all variables in parallel.
problem Sampling complex high-dimensional discrete distributions efficiently and accurately.
method Discrete Langevin proposal (DLP) for parallel coordinate updates with controlled stepsize.
result DLP efficiently explores high-dimensional and strongly correlated variables with asymptotic bias of zero for log-quadratic distributions.
A scalable parallel BO method for asynchronous settings.
problem Expensive-to-evaluate problems in machine learning.
method Simple and scalable Bayesian optimization method for asynchronous parallel settings.
result Demonstrated promising performance on benchmark functions and hyperparameter optimization.
Stochastic gradient descent (SGD) is a popular stochastic optimization method in machine learning. Traditional parallel SGD algorithms, e.g., SimuParallel SGD, often require all nodes to have the same performance or to consume equal quantities of data. However, these requirements are difficult to satisfy when the paral…
Mango automates hyperparameter tuning for large-scale ML training.
problem Manual hyperparameter tuning is tedious and inefficient for large-scale machine learning.
method Parallel hyperparameter tuning with intelligent search strategies and flexible abstractions.
result Mango achieves comparable performance to Hyperopt while supporting distributed computing.
This paper tackles learning functions on manifolds using parallel distributed learning.
problem Learning real-valued functions on manifolds from input-output data pairs.
method Filtered hyperinterpolation and parallel distributed learning.
result Optimal approximation order for non-distributed case, and quantitative relations for distributed case.
Deep learning is a popular machine learning technique and has been applied to many real-world problems. However, training a deep neural network is very time-consuming, especially on big data. It has become difficult for a single machine to train a large model over large datasets. A popular solution is to distribute and…
Real AlphaBeta-Geometriesmath.DG
By a real alphabeta-geometry we mean a four-dimensional manifold M equipped with a neutral metric h such that (M,h) admits both an integrable distribution of alpha-planes and an integrable distribution of beta-planes. We obtain a local characterization of the metric when at least one of the distributions is parallel (i…
A new framework for efficient Bayesian network inference.
problem High-dimensional Bayesian networks are hard to infer due to computational scaling.
method Directed convex subgraphs and minimal d-decomposition tree for decomposition, enabling parallel computation.
result The method reduces computational cost and enables parallel computation.
Benefitting from large-scale training datasets and the complex training network, Convolutional Neural Networks (CNNs) are widely applied in various fields with high accuracy. However, the training process of CNNs is very time-consuming, where large amounts of training samples and iterative operations are required to ob…
The paper proves a convergence theorem for Wiener measures on holonomy groups.
problem Understanding convergence of Wiener measures on holonomy groups.
method Using stochastic parallel transports along convergent metric connections.
result Proves a convergence theorem for push-forward Wiener measures on holonomy groups.
Improved DOA estimation with distributed sensors across multiple frequencies.
problem Sensor gain uncertainties and directional perturbations in multi-frequency scenarios.
method Distributed optimization with local coherence models and iterative exchange of information.
result Advantages in statistical and computational efficiency through parallel iterative technique.