arXiv research

New decision-theoretic characterization separates belief and decision posteriors.

Bayesian approach to selecting data for machine learning.

Bayes classifier cannot be learned from noisy labels without knowing noise distribution.

New online GP algorithm offers performance guarantees for streaming data.

This paper solves the open problem of computing Bayes optimal prediction for decision trees using a Markov chain Monte Carlo method.

This paper explores minimax-Bayes solutions for reinforcement learning problems.

This paper proposes an efficient method for calculating Shapley values in Naive Bayes classifiers.

This work bounds classification error in machine learning for low Bayes error conditions.

A new sequential method estimates Poisson means in streaming data, achieving optimality and efficiency.

Paper extends transfer learning for decision rules, improving treatment rule estimation.

Bayesian framework for policy learning in decision problems.

Survey and compare PAC-Bayes bounds for bandit problems.

Optimizes portfolio construction using Bayesian methods and variational techniques.

This paper establishes the asymptotic consistency of the {\it loss-calibrated variational Bayes} (LCVB) method. LCVB was proposed in~\cite{LaSiGh2011} as a method for approximately computing Bayesian posteriors in a `loss aware' manner. This methodology is also highly relevant in general data-driven decision-making con…

The paper explores robust classifiers for imbalanced Gaussian data.

Bayesian framework reduces online optimization regret.

A new procedure for learning cost-sensitive SVM(CS-SVM) classifiers is proposed. The SVM hinge loss is extended to the cost sensitive setting, and the CS-SVM is derived as the minimizer of the associated risk. The extension of the hinge loss draws on recent connections between risk minimization and probability elicitat…

New loss improves DNN calibration without sacrificing accuracy.

This article introduces a framework to estimate the value of evidence-based decision making.

The NL score optimizes speaker recognition tasks.

A new probability distribution on full rooted trees helps in model selection.

Test-time training adapts a pretrained model to each prompt via parameter updates, improving accuracy under pretraining-to-test distribution shifts.

Improves decision-making in models fit with AEVB by using distinct approximate posteriors.

This study's goal is to create a model of sentiment analysis on a 2000 rows IMDB movie comments and 3200 Twitter data by using machine learning and vector space techniques; positive or negative preliminary information about the text is to provide. In the study, a vector space was created in the KNIME Analytics platform…

Bayesian principles improve agentic AI decision-making.

Study optimizes best-arm identification with minimax and Bayes strategies.

DCC separates marginal estimation from dependence modeling for improved classification accuracy.

Study on limits of LLM-based multi-agent planning reliability.

We show how to compute lower bounds for the supremum Bayes error if the class-conditional distributions must satisfy moment constraints, where the supremum is with respect to the unknown class-conditional distributions. Our approach makes use of Curto and Fialkow's solutions for the truncated moment problem. The lower …

The developments of Rademacher complexity and PAC-Bayesian theory have been largely independent. One exception is the PAC-Bayes theorem of Kakade, Sridharan, and Tewari (2008), which is established via Rademacher complexity theory by viewing Gibbs classifiers as linear operators. The goal of this paper is to extend thi…

We show how to compute the Bayes error-rate for speaker verifiers.

Meta-learning bounds derived using PAC-Bayes theory for improved generalization.

We present the first PAC optimal algorithm for Bayes-Adaptive Markov Decision Processes (BAMDPs) in continuous state and action spaces, to the best of our knowledge. The BAMDP framework elegantly addresses model uncertainty by incorporating Bayesian belief updates into long-term expected return. However, computing an e…

Bayesian neural networks are shown to be minimax and admissible under certain conditions.

Bayes-TrEx finds in-distribution examples for model inspection.

Introduces PAC-Bayes bounds for understanding learning procedures.

Paper develops a new generalization bound using PAC-Bayes theory and Gibbs distributions.

Boosting meta-trees improve decision tree performance.

We describe two techniques that significantly improve the running time of several standard machine-learning algorithms when data is sparse. The first technique is an algorithm that effeciently extracts one-way and two-way counts--either real or expected-- from discrete data. Extracting such counts is a fundamental step…

Reward augmented maximum likelihood (RAML), a simple and effective learning framework to directly optimize towards the reward function in structured prediction tasks, has led to a number of impressive empirical successes. RAML incorporates task-specific reward by performing maximum-likelihood updates on candidate outpu…

Unified framework for Bayes-optimal classifiers under group fairness.

Deep learning improves Bayes factor computation for likelihood-free models.

PAC-Bayes bounds for Gibbs posteriors derived via singular learning theory.

New method reduces computational cost for estimating PAC-Bayes bounds.

A new method estimates Bayes error for deep networks, suggesting they may have reached the limit.

qEUBO optimizes decision-making with noisy feedback.

New PAC-Bayes bounds derived using Legendre transform and f-divergences.

As machine learning is increasingly used to make real-world decisions, recent research efforts aim to define and ensure fairness in algorithmic decision making. Existing methods often assume a fixed set of observable features to define individuals, but lack a discussion of certain features not being observed at test ti…