arXiv research

A new restart criterion for k-means++ improves clustering quality and adapts to data difficulty.

Recent studies have shown that proximal gradient (PG) method and accelerated gradient method (APG) with restarting can enjoy a linear convergence under a weaker condition than strong convexity, namely a quadratic growth condition (QGC). However, the faster convergence of restarting APG method relies on the potentially …

We consider the generic approach of using an experience memory to help exploration by adapting a restart distribution. That is, given the capacity to reset the state with those corresponding to the agent's past observations, we help exploration by promoting faster state-space coverage via restarting the agent from a mo…

Study adapts combinatorial semi-bandit for piecewise stationary, causally related rewards.

Paper adds a restart mechanism to a drawdown control policy for better trading performance.

New method tracks significant arm switches to improve bandit algorithms.

A new sampling method called Restart improves both speed and quality of generative processes.

BRPC online Bayesian calibration handles gradual and abrupt system changes.

New algorithms boost SAT solver performance by optimizing restart strategies.

Master algorithm fails to detect non-stationarity in practical settings.

Motivated by their broad applications in reinforcement learning, we study the linear two-time-scale stochastic approximation, an iterative method using two different step sizes for finding the solutions of a system of two equations. Our main focus is to characterize the finite-time complexity of this method under time-…

Improved SEG method converges to Nash equilibrium in bilinear games.

A new algorithm solves nonnegative least squares faster with nonnegative data.

RePS improves diffusion models for solving inverse problems efficiently.

There is a general trend towards solving problems suited to deep learning with more complex deep learning architectures trained on larger training sets. This requires longer compute times and greater data parallelization or model parallelization. Both data and model parallelism have been historically faster in paramete…

State-of-the-art adversarial attacks on neural networks use expensive iterative methods and numerous random restarts from different initial points. Iterative FGSM-based methods without restarts trade off performance for computational efficiency because they do not adequately explore the image space and are highly sensi…

We derive an optimal policy for adaptively restarting a randomized algorithm, based on observed features of the run-so-far, so as to minimize the expected time required for the algorithm to successfully terminate. Given a suitable Bayesian prior, this result can be used to select the optimal black-box optimization algo…

New RL algorithm tackles nonstationary MDPs with linear approximations and varying rewards.

Algorithm adapts to non-stationary rewards without prior knowledge.

WARPd method solves inverse problems with approximate sharpness conditions.

In this paper we develop a Morse theory for the uniform energy. We use the one-sided directional derivative of the distance function to study the minimizing properties of variations through closed geodesics. This derivative is then used to define a one-sided directional derivative for the uniform energy which allows us…

PER-ETD improves ETD by reducing variance to polynomial complexity.

A parameter-free PGD algorithm for convex optimization.

We develop a projected Nesterov's proximal-gradient (PNPG) approach for sparse signal reconstruction that combines adaptive step size with Nesterov's momentum acceleration. The objective function that we wish to minimize is the sum of a convex differentiable data-fidelity (negative log-likelihood (NLL)) term and a conv…

Online distributional prediction with latent cluster geometry

New bandit algorithm detects and adapts to seasonal changes in rewards.

We consider the problem of online forecasting of sequences of length $n$ with total-variation at most $C_n$ using observations contaminated by independent $σ$-subgaussian noise. We design an $O(n\log n)$-time algorithm that achieves a cumulative square error of $\tilde{O}(n^{1/3}C_n^{2/3}σ^{4/3} + C_n^2)$ with high pro…

LLMs optimize quantum circuits by iteratively improving proposals with feedback and memory traces.

Simple adaptive attacks can jailbreak state-of-the-art safety-aligned language models.

Paper tackles non-stationary kernelized bandits with near-optimal algorithm.

Continuous optimization is an important problem in many areas of AI, including vision, robotics, probabilistic inference, and machine learning. Unfortunately, most real-world optimization problems are nonconvex, causing standard convex techniques to find only local optima, even with extensions like random restarts and …

ParaMonte::Python streamlines Bayesian data analysis with fast Monte Carlo and MCMC routines.

New algorithms improve boosting by optimizing chance-corrected measures.

PROPO tackles non-stationary MDPs with efficient policy optimization.

The convergence rate and final performance of common deep learning models have significantly benefited from heuristics such as learning rate schedules, knowledge distillation, skip connections, and normalization layers. In the absence of theoretical underpinnings, controlled experiments aimed at explaining these strate…

One-pass optimisation for high-dimensional hyperparameters.

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…

New algorithm optimizes beam and rate allocation in mmWave systems for multiple users.

This paper is concerned with improving the empirical convergence speed of block-coordinate descent algorithms for approximate nonnegative tensor factorization (NTF). We propose an extrapolation strategy in-between block updates, referred to as heuristic extrapolation with restarts (HER). HER significantly accelerates t…

Dual training method for EBMs with overparametrized neural networks.

Unified approach for non-stationary linear bandits with dynamic regret.

New RL algorithm tackles non-stationary environments with flexible policy updates.

Algorithm minimizes control regret for non-stationary LQR systems.

In this paper, we propose a distributed algorithm for stochastic smooth, non-convex optimization. We assume a worker-server architecture where $N$ nodes, each having $n$ (potentially infinite) number of samples, collaborate with the help of a central server to perform the optimization task. The global objective is to m…

Determinantal consensus clustering improves clustering robustness.

Multi-objective Neural Architecture Search (NAS) aims to discover novel architectures in the presence of multiple conflicting objectives. Despite recent progress, the problem of approximating the full Pareto front accurately and efficiently remains challenging. In this work, we explore the novel reinforcement learning …

Algorithm reduces decision-making errors in multi-agent bandit problems.

SRSGD improves DNN training speed and accuracy.