The 28th International Conference on Artificial Intelligence and Statistics

The 28th International Conference on Artificial Intelligence and Statistics (AISTATS 2025) is one of the most selective and respected venues at the intersection of machine learning and statistical theory, dedicated to research that advances both fields simultaneously. Topics include probabilistic models, Bayesian inference, causal discovery, variational methods, statistical learning theory, deep learning foundations, and the development of rigorous evaluation frameworks that connect empirical results to theoretical guarantees.

AISTATS attracts machine learning researchers, statisticians, and applied mathematicians who value theoretical rigor alongside empirical relevance, and who recognize that the most durable advances in AI require principled statistical foundations. The conference's small size and focused scope create an unusually high concentration of expertise, making it a prime venue for graduate students and established researchers alike to engage directly with the theoreticians shaping the mathematical underpinnings of modern machine learning.