arXiv research

Study agnostic RL in large state spaces with weak function approximation.

Improved agnostic learning time via Gaussian surface area analysis.

New insights into RL with weak function approximation.

First proper learning algorithm for Gaussian halfspaces with matching sample and computational complexity.

New algorithm finds optimal policy with polynomial trajectories in deterministic systems.

Efficiently learns Single-Index Models with constant factor approximation.

We obtain the first positive results for bounded sample compression in the agnostic regression setting with the $\ell_p$ loss, where $p\in [1,\infty]$. We construct a generic approximate sample compression scheme for real-valued function classes exhibiting exponential size in the fat-shattering dimension but independen…

The study optimizes polynomial regression for learning under Gaussian distributions.

New SQ lower bound shows complexity nearly matches known upper bound for smoothed agnostic learning.

Robust learner finds subspace for MIMs with label noise.

Improved agnostic boosting with better sample efficiency.

Study shows depth improves generalization in deep learning models.

New algorithm learns halfspaces over hypercube with random bit flips.

Paper analyzes error exponent in agnostic PAC learning.

Study selective classification with halfspaces, achieving error bounds under Gaussian distributions.

Study shows transductive learning is equivalent to PAC learning for most natural loss functions.

Learning to infer Bayesian posterior from a few-shot dataset is an important step towards robust meta-learning due to the model uncertainty inherent in the problem. In this paper, we propose a novel Bayesian model-agnostic meta-learning method. The proposed method combines scalable gradient-based meta-learning with non…

Study learning and refutation in non-interactive LDP, showing sample complexity equivalence.

This paper introduces the probabilistic module interface, which allows encapsulation of complex probabilistic models with latent variables alongside custom stochastic approximate inference machinery, and provides a platform-agnostic abstraction barrier separating the model internals from the host probabilistic inferenc…

Develops a transparent surrogate model for complex data.

A new method learns fair classifiers without sacrificing accuracy.

A new method reduces the computational cost of KernelSHAP for explaining predictions.

We study tensor completion in the agnostic setting. In the classical tensor completion problem, we receive $n$ entries of an unknown rank-$r$ tensor and wish to exactly complete the remaining entries. In agnostic tensor completion, we make no assumption on the rank of the unknown tensor, but attempt to predict unknown …

We introduce a new model-agnostic explanation technique which explains the prediction of any classifier called CLE. CLE gives an faithful and interpretable explanation to the prediction, by approximating the model locally using an interpretable model. We demonstrate the flexibility of CLE by explaining different models…

A new framework for brain mapping using statistical agnostic methods.

A new method approximates posterior for VAEs without iterative training.

New research shows logistic regression can achieve optimal error rate for agnostic learning of halfspaces.

Extends boosting to multiclass online agnostic classification.

We establish a tight characterization of the worst-case rates for the excess risk of agnostic learning with sample compression schemes and for uniform convergence for agnostic sample compression schemes. In particular, we find that the optimal rates of convergence for size-$k$ agnostic sample compression schemes are of…

Our research proves neural collapse in deep ResNets and transformers is globally optimal.

MAML optimizes shared priors for subtasks in a nonconvex meta-objective.

New method handles unknown task boundaries in continual learning.

The paper corrects Bayesian neural network approximations to improve decision quality.

New algorithm learns disjunctions faster than previous methods.

New method for private density estimation of high-dimensional Gaussian mixtures.

LUNO linearizes neural operators to quantify their predictive uncertainty.

New methods for fairness in regression using probabilistic classification.

Building an open-domain conversational agent is a challenging problem. Current evaluation methods, mostly post-hoc judgments of static conversation, do not capture conversation quality in a realistic interactive context. In this paper, we investigate interactive human evaluation and provide evidence for its necessity; …

Recent work in model-agnostic explanations of black-box machine learning has demonstrated that interpretability of complex models does not have to come at the cost of accuracy or model flexibility. However, it is not clear what kind of explanations, such as linear models, decision trees, and rule lists, are the appropr…

New algorithms improve agnostic learning for triangles and polygons, reducing time complexity.

New algorithms save computation in agnostic learning with membership queries.

New algorithms improve label complexity for active multi-distribution learning.

We consider the problem of estimating the mean and covariance of a distribution from iid samples in $\mathbb{R}^n$, in the presence of an $η$ fraction of malicious noise; this is in contrast to much recent work where the noise itself is assumed to be from a distribution of known type. The agnostic problem includes many…

We revisit the problem of differentially private release of classification queries. In this problem, the goal is to design an algorithm that can accurately answer a sequence of classification queries based on a private training set while ensuring differential privacy. We formally study this problem in the agnostic PAC …

New bounds for agnostic learning with average smoothness.

Query access significantly speeds up learning Multi-Index Models under Gaussian distribution.

Boosting with unlabeled data achieves optimal sample complexity in agnostic settings.

Proposes class-agnostic object detection to handle all objects without class labels.