arXiv research

Gradient methods converge better for alternating updates in bilinear zero-sum games.

Games generalize the single-objective optimization paradigm by introducing different objective functions for different players. Differentiable games often proceed by simultaneous or alternating gradient updates. In machine learning, games are gaining new importance through formulations like generative adversarial netwo…

Alt-GDA outperforms Sim-GDA in minimax games with near-optimal local convergence.

New method for clustering and optimization handles evolving datasets efficiently.

Bayesian model updating uses VAEs to approximate likelihood with small data.

We show that a simple model of a spatially resolved evolving economic system, which has a steady state under simultaneous updating, shows stable oscillations in price when updated asynchronously. The oscillations arise from a gradual decline of the mean price due to competition among sellers competing for the same reso…

Recently, a lot of effort has been paid to the efficient computation of Kriging predictors when observations are assimilated sequentially. In particular, Kriging update formulae enabling significant computational savings were derived in Barnes and Watson (1992), Gao et al. (1996), and Emery (2009). Taking advantage of …

Two algorithms for linear contextual bandits with rare updates achieve optimal regret and efficiency.

Paper addresses OPE for dependent bandit samples using MDS and batch updates.

Proposes a method to balance tasks in multitask learning with a single gradient step update.

Regression problems that have closed-form solutions are well understood and can be easily implemented when the dataset is small enough to be all loaded into the RAM. Challenges arise when data is too big to be stored in RAM to compute the closed form solutions. Many techniques were proposed to overcome or alleviate the…

Being able to reach any desired location in the environment can be a valuable asset for an agent. Learning a policy to navigate between all pairs of states individually is often not feasible. An all-goals updating algorithm uses each transition to learn Q-values towards all goals simultaneously and off-policy. However …

Simultaneous Latent Budget Trees for stratified classification

A new FL algorithm reduces communication overhead by selectively updating model parameters.

Proposes glocal hypergradient estimation for hyperparameter optimization.

Two algorithms achieve optimal regret with limited adaptivity in multinomial logistic bandits.

Proposes a new method for finding frequent closed patterns in transaction bases.

Federated learning distributes model training among a multitude of agents, who, guided by privacy concerns, perform training using their local data but share only model parameter updates, for iterative aggregation at the server. In this work, we explore the threat of model poisoning attacks on federated learning initia…

M-flows learn data manifolds and densities, improving manifold learning and inference.

We analyze SA with Markovian data and nonlinear updates, overcoming prior limitations.

Single-timescale actor-critic finds globally optimal policy.

BC-Aligner maintains backward compatibility of embeddings after frequent updates.

FedElasticNet reduces communication costs and handles client drift in FL.

The paper simplifies multi-agent RL dynamics in finite-state Markov games using homogenization.

AL modularizes deep learning networks, improving training speed and accuracy.

Mathematical methods characterize RNNs' asymptotics as hidden units and data grow.

A critical and challenging problem in reinforcement learning is how to learn the state-action value function from the experience replay buffer and simultaneously keep sample efficiency and faster convergence to a high quality solution. In prior works, transitions are uniformly sampled at random from the replay buffer o…

New algorithm reduces FL sample and communication costs.

NVA combines variational posteriors, annealing, and natural-gradient learning for multimodal optimization.

Jointly reduces echo, reverberation, and noise using neural networks.

Consider convex optimization problems subject to a large number of constraints. We focus on stochastic problems in which the objective takes the form of expected values and the feasible set is the intersection of a large number of convex sets. We propose a class of algorithms that perform both stochastic gradient desce…

Study allocates resources to strategic agents while balancing cost and incentives.

EP learns like BPTT but with local weight updates.

Matrix completion, where we wish to recover a low rank matrix by observing a few entries from it, is a widely studied problem in both theory and practice with wide applications. Most of the provable algorithms so far on this problem have been restricted to the offline setting where they provide an estimate of the unkno…

DG separates successes and failures by gating updates with advantage and surprisal.

Proposes dynamic channel pruning during neural network training.

We study the sparse non-negative least squares (S-NNLS) problem. S-NNLS occurs naturally in a wide variety of applications where an unknown, non-negative quantity must be recovered from linear measurements. We present a unified framework for S-NNLS based on a rectified power exponential scale mixture prior on the spars…

Gradient-based methods for games suffer from discrete update steps that cause drift, affecting performance.

The group lasso is a penalized regression method, used in regression problems where the covariates are partitioned into groups to promote sparsity at the group level. Existing methods for finding the group lasso estimator either use gradient projection methods to update the entire coefficient vector simultaneously at e…

We propose a novel approach to model selection for simulator-based statistical models. The proposed approach defines a mixture of candidate models, and then iteratively updates the weight coefficients for those models as well as the parameters in each model simultaneously; this is done by recursively applying Bayes' ru…

MWGraD solves multi-objective distributional optimization using particle-based gradient descent.

Deep learning model improves financial return forecasting using LOBs.

Paper introduces REED for noncoherent OTA-FL, reducing latency without phase alignment.

Unified proof for scalable personalized federated learning.

Enhances neural architecture search efficiency and prevents performance collapse.

End-to-end graph-based SSL learns all graph factors dynamically.

Proposes CRG_IMSC for better clustering of multi-view data.

Despite the recent popularity of deep generative state space models, few comparisons have been made between network architectures and the inference steps of the Bayesian filtering framework -- with most models simultaneously approximating both state transition and update steps with a single recurrent neural network (RN…