AGNES accelerates gradient descent with noisy gradients.
problem Minimizing smooth convex and strongly convex functions with noisy gradients.
method Generalization of Nesterov's accelerated gradient descent algorithm for noisy conditions.
result AGNES achieves acceleration for noisy gradients with a constant of proportionality up to 1.
New method accelerates CNNs for mobile devices by approximating tensors and quantizing weights.
problem Efficiently compress and accelerate CNNs for mobile devices.
method Low-rank tensor approximation in Tucker format combined with quantization of weights and activations.
result Our method significantly improves CNN performance on various classification tasks.
Accelerates TD learning for long-horizon reinforcement learning problems.
problem Slow convergence of conventional TD learning in long-horizon tasks.
method Introduces PID Accelerated Temporal Difference (PID TD) learning algorithms.
result Accelerates convergence of TD learning compared to conventional methods.
PTF accelerates RL by reusing source policies without measuring task similarity.
problem Leveraging prior knowledge for faster RL.
method Adaptive Policy Transfer Framework (PTF) for RL.
result Significantly accelerates RL learning process and surpasses state-of-the-art methods.
Memory-augmented neural networks (MANNs) are designed for question-answering tasks. It is difficult to run a MANN effectively on accelerators designed for other neural networks (NNs), in particular on mobile devices, because MANNs require recurrent data paths and various types of operations related to external memory a…
Weak labels can significantly speed up learning for strong tasks.
problem Learning with limited strong labels.
method Using weak labels to accelerate learning of strong tasks.
result Weak labels can accelerate learning to O ( i c e f r a c 1 n ) \mathcal{O}(
icefrac{1}{n}) O ( i ce f r a c 1 n ) rate. DeltaBO accelerates Bayesian optimization with theoretical guarantees.
problem Improving Bayesian optimization performance through knowledge transfer.
method DeltaBO algorithm that transfers knowledge from related source tasks to target tasks using the difference function.
result DeltaBO achieves a regret bound of O ~ ( T ( T / N + γ δ ) ) \widetilde{O}(\sqrt{T (T/N + γ_δ)}) O ( T ( T / N + γ δ ) ) , significantly better than existing methods. Super-acceleration of gradient descent with momentum improves loss function minimization.
problem Minimizing loss functions in machine learning.
method Extending Nesterov acceleration by using gradients at multiple steps ahead.
result Super-acceleration of the momentum algorithm is beneficial for various loss landscapes and tasks.
Accelerates MMLE using SVGD with Nesterov acceleration.
problem Maximum Marginal Likelihood Estimation optimization.
method Stein variational gradient descent with Nesterov acceleration.
result Consistently accelerates convergence across various tasks.
Model-free deep reinforcement learning (RL) algorithms have been widely used for a range of complex control tasks. However, slow convergence and sample inefficiency remain challenging problems in RL, especially when handling continuous and high-dimensional state spaces. To tackle this problem, we propose a general acce…
Accelerates sampling from Gibbs distributions using ARWP method.
problem Sampling from Gibbs distributions efficiently.
method ARWP method, combining Nesterov acceleration and regularized Wasserstein proximal.
result ARWP exhibits higher contraction rate and faster tail exploration.
Derives PF-ODE for infinite-dimensional functions, improving function generation tasks.
problem Efficient inference in infinite-dimensional diffusion models.
method Derives PF-ODE in infinite-dimensional function spaces.
result Reduces function evaluations while maintaining sample quality.
Accelerated RPCholesky speeds up kernel matrix approximations.
problem Efficiently approximating large kernel matrices.
method Accelerated randomly pivoted Cholesky (RPCholesky) with block matrix computations and rejection sampling.
result Approximates kernel matrices up to 40 times faster.
SuNCEt accelerates contrastive learning with minimal labeled data.
problem Efficiently learning visual representations with limited labeled data.
method Noise-contrastive estimation and neighbourhood component analysis-based semi-supervised loss.
result SuNCEt achieves semi-supervised learning accuracy with less than half the labeled data.
SDM Policy accelerates inference for robotic tasks while maintaining high action quality.
problem Prolonged inference times in diffusion-based policies for high-frequency control tasks.
method Two-stage optimization: score matching and distribution matching; dual-teacher mechanism.
result 6x inference speedup with state-of-the-art action quality.
Paper tackles drowsy driving by learning from weakly labeled car acceleration data.
problem Lack of labeled data for estimating driver drowsiness.
method Weakly supervised learning, scalable stochastic optimization.
result Algorithm learns from weakly labeled data, outperforming baseline methods.
Synthetic data can be used to ask more questions and accelerate discovery with provable validity guarantees.
problem Valid inference with synthetic data
method Task exchangeability
result Provable validity guarantees for synthetic data inference
New algorithm accelerates optimization on Riemannian manifolds, including Wasserstein space.
problem Accelerating optimization methods in Riemannian geometry.
method Dynamic stepsize algorithms on Riemannian manifolds with specific vector transport.
result First provable accelerated gradient method in Wasserstein space.
A deep-learning inference accelerator is synthesized from a C-language software program parallelized with Pthreads. The software implementation uses the well-known producer/consumer model with parallel threads interconnected by FIFO queues. The LegUp high-level synthesis (HLS) tool synthesizes threads into parallel FPG…
New method speeds up lifelong learning of complex tasks.
problem Catastrophic forgetting in lifelong learning.
method Directly trains lifelong function approximators via policy gradients.
result Faster convergence and better policies achieved.
The Regularized Nonlinear Acceleration (RNA) algorithm is an acceleration method capable of improving the rate of convergence of many optimization schemes such as gradient descend, SAGA or SVRG. Until now, its analysis is limited to convex problems, but empirical observations shows that RNA may be extended to wider set…
New approach uses unlabeled prior data to accelerate exploration in sparse reward tasks.
problem Sparse reward tasks in reinforcement learning.
method Learn reward model from online experience, label prior data, and use concurrently.
result Rapid exploration in challenging sparse-reward domains.
Improved NiNo networks accelerate Adam training by up to 50%.
problem Accelerating neural network training with stable and efficient methods.
method Proposed NiNo networks that leverage neuron connectivity and graph neural networks to nowcast parameters periodically during Adam training.
result Accelerates Adam training by up to 50% in vision and language tasks.
A Python package for GPU-accelerated signature kernel computation.
problem Efficient computation of signature kernels for sequential data.
method GPU-accelerated algorithms and tensor sketches.
result New algorithm outperforms existing methods.
Accelerates policy optimization in RL with optimistic and adaptive updates.
problem Improving policy optimization methods in reinforcement learning.
method Integrates foresight into policy improvement step via optimistic and adaptive updates.
result Designs an optimistic policy gradient algorithm, adaptive via meta-gradient learning.
Momentum accelerates Frank Wolfe algorithms on certain problems.
problem Improving convergence rate of Frank Wolfe algorithms.
method Introducing momentum into Frank Wolfe algorithms and proving faster convergence rate.
result Accelerated Frank Wolfe (AFW) converges with a faster rate of i l d e O ( 1 k 2 ) ilde{\cal O}(\frac{1}{k^2}) i l d e O ( k 2 1 ) . Hardware-accelerated RBM solves large combinatorial problems and integer factorization.
problem Solving large combinatorial optimization and integer factorization problems.
method Logically synthesized RBM architecture, hardware acceleration, and efficient training methods.
result Hardware-accelerated RBM factorizes 16-bit numbers with 10000x speed and 32x power improvements.
SUOD accelerates OD for large, diverse models.
problem Training and scoring new samples with many unsupervised, heterogeneous OD models.
method Data reduction, model approximation, and taskload optimization.
result SUOD accelerates OD for over 20 benchmark datasets and a real-world case.
In recent years, deep neural networks (DNN) have demonstrated significant business impact in large scale analysis and classification tasks such as speech recognition, visual object detection, pattern extraction, etc. Training of large DNNs, however, is universally considered as time consuming and computationally intens…
Improved SDE-BNN model reduces NFEs and accelerates convergence.
problem High computational cost and convergence instability in SDE-BNNs.
method Nesterov's Accelerated Gradient (NAG) method integrated into SDE-BNN framework.
result Significantly reduced number of function evaluations (NFEs) and improved predictive accuracy.
Study uses multi-task Bayesian optimization to speed up SVM hyperparameter tuning for nodules diagnosis.
problem Redundant and time-consuming hyperparameter tuning for SVM classifiers in medical imaging.
method Employed multi-task Bayesian optimization to accelerate hyperparameter search.
result Multi-task Bayesian optimization significantly accelerates hyperparameter search.
New algorithms accelerate SVGD convergence using deep unfolding.
problem Improving the speed of SVGD convergence.
method Integrating deep unfolding into SVGD for parameter learning.
result Proposed algorithms achieve faster convergence in various tasks.
In many real-world scenarios, an autonomous agent often encounters various tasks within a single complex environment. We propose to build a graph abstraction over the environment structure to accelerate the learning of these tasks. Here, nodes are important points of interest (pivotal states) and edges represent feasib…
Paper introduces a meta-critic for accelerating off-policy actor-critic learning.
problem Improving sample efficiency in continuous control tasks.
method Meta-critic that meta-learns an additional loss for the actor.
result Online meta-critic learning leads to improved performance in various continuous control environments.
This paper introduces an acceleration structure for hyperbolic embeddings.
problem Efficiently embedding and visualizing high-dimensional data in hyperbolic spaces.
method Building upon a polar quadtree, the paper introduces a new acceleration structure for hyperbolic embeddings.
result The new method computes embeddings in significantly less time compared to existing methods.
SVRN accelerates Newton methods by reducing variance and improving performance.
problem Improving the efficiency of Newton methods for large-scale optimization problems.
method Stochastic Variance-Reduced Newton (SVRN) algorithm that accelerates Subsampled Newton and Iterative Hessian Sketch algorithms.
result SVRN accelerates Newton methods by reducing the number of passes over the data, achieving a significant improvement in performance.
Graph neural networks speed up nonnegative matrix factorization.
problem Efficiently factorize nonnegative matrices for various applications.
method Developed a graph neural network that combines bipartite self-attention with ADMM updates.
result Significant acceleration achieved in nonnegative matrix factorization.
DeepLight accelerates CTR predictions in ad serving by 46X.
problem Significantly increased serving delay and high memory usage for ad serving.
method Explicitly searching feature interactions, pruning layers, promoting sparsity.
result Accelerates model inference by 46X on Criteo dataset.
In this paper, we propose a novel progressive parameter pruning method for Convolutional Neural Network acceleration, named Structured Probabilistic Pruning (SPP), which effectively prunes weights of convolutional layers in a probabilistic manner. Unlike existing deterministic pruning approaches, where unimportant weig…
Proposes a method to accelerate safe sequential learning using offline data.
problem Limited exploration due to disconnected safe regions and slow task learning.
method Safe transfer sequential learning using Gaussian processes and offline data.
result Enhances global exploration across multiple disjoint safe regions with lower data consumption.
HAL accelerates the generation of training sets for accurate interatomic potentials.
problem Generating accurate and transferable interatomic potentials is time-consuming and requires expert input.
method HAL framework using a physically motivated sampler with a biasing term to drive high uncertainty configurations.
result HAL-generated training databases for alloys and polymers predict macroscopic properties with high accuracy.
ES improves training efficiency by dynamically selecting data samples.
problem Efficiently selecting informative data samples for faster learning.
method Evolved Sampling (ES) dynamically selects data samples based on loss dynamics and differences.
result ES achieves significant training acceleration without compromising model performance.
Unified framework for accelerating DNNs on resource-limited platforms.
problem Accelerating DNN execution on resource-limited platforms.
method Block-based pruning framework with reweighted regularization.
result First universal framework for both CNNs and RNNs with real-time acceleration and no accuracy compromise.
We present a framework for accelerating a spectrum of machine learning algorithms that require computation of bilinear inverse forms u ⊤ A − 1 u u^\top A^{-1}u u ⊤ A − 1 u , where A A A is a positive definite matrix and u u u a given vector. Our framework is built on Gauss-type quadrature and easily scales to large, sparse matrices. Further, it …
Paper proposes NASAIC framework for co-designing neural architectures and heterogeneous ASICs.
problem Designing efficient neural architectures and ASICs for multiple tasks.
method Build ASIC templates and propose NASAIC framework for simultaneous design of architectures and ASICs.
result NASAIC ensures design specifications and maximizes accuracy with minimal performance loss.
Parallel score matching accelerates DPM training and improves density estimation.
problem Extended training periods and limited modeling flexibility in DPMs.
method Partitioning the learning task into independent time sub-intervals and modeling the score at each time point separately.
result Significant acceleration of training process and improved density estimation performance.
We accelerate CNF by reducing ODE truncation errors with polynomial regularization.
problem High computation cost of CNF due to large truncation errors in solving ODEs.
method Add polynomial regularization to approximate ODE trajectories with polynomial functions.
result 42.3% to 71.3% reduction of NFE on density estimation, 19.3% to 32.1% on variational auto-encoder.
This paper accelerates D-PSGD and AD-PSGD for large-scale deep learning tasks.
problem Decreasing spectral gap with increasing number of learners hampers convergence in D-PSGD and AD-PSGD.
method Improves spectral gap while minimizing communication cost through new techniques.
result Demonstrates faster training times and lower error rates on large-scale tasks.