arXiv research

New framework allows reinforcement learning with polynomial sample complexity.

New method learns near-optimal policies with polynomial samples in A and H.

Paper tackles efficient IRL in offline settings with polynomial samples and runtime.

New method learns to generalize across different data domains efficiently.

New RL algorithms achieve optimal policies with polynomial sample complexity for mean-field problems.

AE-LSVI identifies near-optimal policies in complex systems with minimal data.

Study shows private learning of mixtures of Gaussians is possible with polynomial samples.

Study analyzes learnability of RKHS under L∞ norm for kernel methods.

New algorithms reduce complexity for learning in MDPs with entropy regularization.

FNOs learn solution operators of dissipative equations efficiently via spectral methods.

New findings on depth vs. width in neural networks, showing depth can improve learnability.

New RL method learns to skip states in linearly $q^π$-realizable MDPs, simplifying to linear MDPs.

New algorithm estimates Gaussian means and covariances efficiently and privately.

This paper advances sample-efficient learning for partially observable RL by introducing B-stability and new algorithms.

We consider the problem of training input-output recurrent neural networks (RNN) for sequence labeling tasks. We propose a novel spectral approach for learning the network parameters. It is based on decomposition of the cross-moment tensor between the output and a non-linear transformation of the input, based on score …

In this paper, we provide a method to learn the directed structure of a Bayesian network using data. The data is accessed by making conditional probability queries to a black-box model. We introduce a notion of simplicity of representation of conditional probability tables for the nodes in the Bayesian network, that we…

Efficient exploration is one of the key challenges for reinforcement learning (RL) algorithms. Most traditional sample efficiency bounds require strategic exploration. Recently many deep RL algorithms with simple heuristic exploration strategies that have few formal guarantees, achieve surprising success in many domain…

New algorithm learns halfspaces with noise using Forster decomposition.

The explore{exploit dilemma is one of the central challenges in Reinforcement Learning (RL). Bayesian RL solves the dilemma by providing the agent with information in the form of a prior distribution over environments; however, full Bayesian planning is intractable. Planning with the mean MDP is a common myopic approxi…

New bounds for score matching in polynomial exponential families.

New findings show limitations in converting private learning to online learning efficiently.

Algorithm finds safe zones in policy Markov Decision Processes to limit trajectory escape.

New framework uses unsupervised learning for efficient exploration in RL.

Training neural networks is a challenging non-convex optimization problem, and backpropagation or gradient descent can get stuck in spurious local optima. We propose a novel algorithm based on tensor decomposition for guaranteed training of two-layer neural networks. We provide risk bounds for our proposed method, with…

Gradient descent learns over-param neural nets better than NTK.

We give a new consistent scoring function for structure learning of Bayesian networks. In contrast to traditional approaches to scorebased structure learning, such as BDeu or MDL, the complexity penalty that we propose is data-dependent and is given by the probability that a conditional independence test correctly show…

The paper solves open questions in computable PAC learning, providing a complete landscape.

In this paper, we consider parameter recovery for non-overlapping convolutional neural networks (CNNs) with multiple kernels. We show that when the inputs follow Gaussian distribution and the sample size is sufficiently large, the squared loss of such CNNs is $\mathit{~locally~strongly~convex}$ in a basin of attraction…

CoT improves transformer sample efficiency by reducing input token dependencies and attention sparsity.

The paper tackles learning mixtures of two multinomial logits, showing identifiability and presenting an algorithm.

We propose a novel parameterized family of Mixed Membership Mallows Models (M4) to account for variability in pairwise comparisons generated by a heterogeneous population of noisy and inconsistent users. M4 models individual preferences as a user-specific probabilistic mixture of shared latent Mallows components. Our k…

ETGL-DDPG improves DDPG for sparse reward control with new exploration and replay techniques.

New algorithms bound graph structure sampling and learning high-dimensional graphical models.

We give a simple, fast algorithm for hyperparameter optimization inspired by techniques from the analysis of Boolean functions. We focus on the high-dimensional regime where the canonical example is training a neural network with a large number of hyperparameters. The algorithm --- an iterative application of compresse…

Study learns a neuron with non-monotonic activation functions.

Paper tackles offline RL with weak assumptions on both function classes and data coverage.

We consider the problem of spherical Gaussian Mixture models with $k \geq 3$ components when the components are well separated. A fundamental previous result established that separation of $Ω(\sqrt{\log k})$ is necessary and sufficient for identifiability of the parameters with polynomial sample complexity (Regev and V…

Modern Reinforcement Learning (RL) is commonly applied to practical problems with an enormous number of states, where function approximation must be deployed to approximate either the value function or the policy. The introduction of function approximation raises a fundamental set of challenges involving computational …

Paper introduces a new reinforcement learning method with improved performance.

In this paper, we propose guaranteed spectral methods for learning a broad range of topic models, which generalize the popular Latent Dirichlet Allocation (LDA). We overcome the limitation of LDA to incorporate arbitrary topic correlations, by assuming that the hidden topic proportions are drawn from a flexible class o…

New algorithm proves RL from partial obs is feasible.

It is becoming increasingly important to understand the vulnerability of machine learning models to adversarial attacks. In this paper we study the feasibility of robust learning from the perspective of computational learning theory, considering both sample and computational complexity. In particular, our definition of…

New algorithms achieve uniform-PAC guarantees for RL with bounded eluder dimension.

Bayesian Reinforcement Learning (RL) is capable of not only incorporating domain knowledge, but also solving the exploration-exploitation dilemma in a natural way. As Bayesian RL is intractable except for special cases, previous work has proposed several approximation methods. However, these methods are usually too sen…

New robust algorithms improve learning with feature feedback.

Paper studies linear CMDPs, improving sample complexity for both models.

New algorithm finds Nash equilibrium in multi-agent games.

Optimistic NPG improves policy optimization in online RL with efficient sample complexity.