Get Started
Home
Topics
Search
Library
Research questionCan stochastic weight averaging improve equivariance in augmented classification without repeated ensemble training?Data augmentation incorporates task symmetries into neural networks, while deep ensembles can require many separate training runs. The practical difficulty is determining whether weight averaging can provide stronger symmetry handling without that repeated cost.
AI
Machine Learning
Neural and Evolutionary Computing
Research Paper
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.Boosting Data Augmentation with Stochastic Weight AveragingThe evidence covers stochastic weight averaging for augmented classification, with an infinite-width Ornstein–Uhlenbeck analysis and numerical experiments across multiple image and graph models with discrete and continuous symmetries.research paper · Sep 4, 2026
Related questions
How can we learn sparse symmetric similarity graphs that are doubly stochastic and enforce a chosen number of clusters?How should multi-expert models balance diversity and prediction variance for long-tailed visual recognition?How can agnostic learning handle low-intrinsic-dimensional concepts under arbitrary distributions with Gaussian-robust benchmarks?How can ensemble classifiers retain accuracy and generalization with far fewer base learners?