Optimal protocols for continual learning via statistical physics and control theory

DOI10.1088/1742-5468/ADF296MaRDI QIDQ7234645zbMATH OpenFDO

Authors Francesco Mori, Stefano Sarao Mannelli, Francesca Mignacco

Publication date 28 April 2026

Published in Journal of Statistical Mechanics: Theory and Experiment (Search for Journal in Brave)

Copyright license Creative Commons Attribution 4.0 International

Full work available at URL https://doi.org/10.1088/1742-5468/ADF296

zbMATH Keywords

learning theorymachine learningonline dynamics

Mathematics Subject Classification ID

Statistical mechanics, structure of matter (82-XX)

Cites work

• A mean field view of the landscape of two-layer neural networks
• A mean-field optimal control formulation of deep learning
• Control theory for physicists
• Disordered high-dimensional optimal control
• Globally optimal on-line learning rules for multi-layer neural networks
• scientific article; zbMATH DE number 1448976 (Why is no real title available?)
• scientific article; zbMATH DE number 7306907 (Why is no real title available?)
• Learning by on-line gradient descent
• Optimization of on-line principal component analysis
• Overcoming catastrophic forgetting in neural networks
• Statistical mechanics of learning
• Trainability and Accuracy of Artificial Neural Networks: An Interacting Particle System Approach

This page was built for publication: Optimal protocols for continual learning via statistical physics and control theory

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7234645)