arXiv research

Mirror descent linked to information ratio via Bayesian regret bounds.

IDS improves sparse linear bandits by balancing information and regret.

Optimizes stochastic linear bandits with efficient, asymptotically optimal algorithm.

New pricing algorithm learns demand curves and optimizes prices in dynamic markets.

New bounds on IDS for RL show how to balance computation and learning efficiency.

Unified algorithm for efficient pure exploration using dual variables.

Bayesian optimization improved for biased data.

Partial monitoring is a rich framework for sequential decision making under uncertainty that generalizes many well known bandit models, including linear, combinatorial and dueling bandits. We introduce information directed sampling (IDS) for stochastic partial monitoring with a linear reward and observation structure. …

We propose a novel algorithm for sequential matrix completion in a recommender system setting, where the $(i,j)$th entry of the matrix corresponds to a user $i$'s rating of product $j$. The objective of the algorithm is to provide a sequential policy for user-product pair recommendation which will yield the highest pos…

ASD algorithm maximizes model estimates by adaptively labeling points.

Efficient exploration remains a major challenge for reinforcement learning. One reason is that the variability of the returns often depends on the current state and action, and is therefore heteroscedastic. Classical exploration strategies such as upper confidence bound algorithms and Thompson sampling fail to appropri…

Novel algorithms for multi-agent reinforcement learning reduce sample complexity.

Hypermodels improve exploration efficiency and accuracy.

IDS improves reinforcement learning with contextual information.

We consider stochastic multi-armed bandit problems with graph feedback, where the decision maker is allowed to observe the neighboring actions of the chosen action. We allow the graph structure to vary with time and consider both deterministic and Erdős-Rényi random graph models. For such a graph feedback model, we fir…

New IDS algorithm refines parameter norm bounds for better bandit performance.

Top-two algorithm improved for best-k-arm selection.

In the stochastic bandit problem, the goal is to maximize an unknown function via a sequence of noisy evaluations. Typically, the observation noise is assumed to be independent of the evaluation point and to satisfy a tail bound uniformly on the domain; a restrictive assumption for many applications. In this work, we c…

A new algorithm STE for model-based RL improves learning rates.

We consider numerical schemes for root finding of noisy responses through generalizing the Probabilistic Bisection Algorithm (PBA) to the more practical context where the sampling distribution is unknown and location-dependent. As in standard PBA, we rely on a knowledge state for the approximate posterior of the root l…

IDS algorithm optimizes sequential decisions in various monitoring settings.

Improved regret bounds for bandit phase retrieval.

Studying a softmax-attention model, we show that the learned query converges to the latent signal subspace spanned by the informative direction.

Improves neural network search in combinatorial spaces of mathematical symbols.

Bayesian metalearning improves performance in linear bandits with misspecified priors.

Information extraction and user intention identification are central topics in modern query understanding and recommendation systems. In this paper, we propose DeepProbe, a generic information-directed interaction framework which is built around an attention-based sequence to sequence (seq2seq) recurrent neural network…

DCGANs generate drainage networks quickly from samples.

Unified framework for ranking-and-selection with multiple correct answers and non-answerable estimates

2L-FUSE enhances feature sparsity through kernel learning.

Two multifidelity trust-region methods use low-fidelity models for efficient optimization.

Proper balance between exploitation and exploration is what makes good decisions, which achieve high rewards like payoff or evolutionary fitness. The Infomax principle postulates that maximization of information directs the function of diverse systems, from living systems to artificial neural networks. While specific a…

Many real-world graphs or networks are temporal, e.g., in a social network persons only interact at specific points in time. This information directs dissemination processes on the network, such as the spread of rumors, fake news, or diseases. However, the current state-of-the-art methods for supervised graph classific…

A new method optimizes projection directions for sliced Wasserstein distances.

Meta-learning performance is affected by how task diversity is allocated, not just overall variability.

Many applications in machine learning require optimizing a function whose true gradient is unknown, but where surrogate gradient information (directions that may be correlated with, but not necessarily identical to, the true gradient) is available instead. This arises when an approximate gradient is easier to compute t…

DDCL-INCRT learns its own structure during training, reducing unnecessary complexity.

Enhanced Sampling Scheme improves masked generative modeling.

PRS improves rejection sampling by learning better proposals.

This paper reviews various sampling methods from statistics and machine learning.

RISA improves VFL by using imputed samples with low uncertainty.

Improved privacy-preserving methods for estimating multiple samples from distributions.

Paper introduces a new sampling method combining Consistency Models with importance sampling.

Neural network accuracy improves with denser training samples.

Wedge Sampling improves tensor completion with nearly-linear sample complexity.

Algorithm samples constrained stochastic differential equations.

In modern data analysis, random sampling is an efficient and widely-used strategy to overcome the computational difficulties brought by large sample size. In previous studies, researchers conducted random sampling which is according to the input data but independent on the response variable, however the response variab…

Generative models map simple samples to complex target samples.

Sampling is a fundamental problem in computer science and statistics. However, for a given task and stream, it is often not possible to choose good sampling probabilities in advance. We derive a general framework for adaptively changing the sampling probabilities via a collection of thresholds.In general, adaptive samp…