00:00:00 / 00:00:00

BI for AI: Energy Conserving Dynamics for optimization and sampling

De Eva Silverstein

Apparaît dans la collection : Mikefest : A conference in honor of Michael Douglas’ 60th birthday

We introduce a novel framework for optimization based on energy-conserving Hamiltonian dynamics in a strongly mixing (chaotic) regime and establish some of its key properties analytically and numerically. The prototype is a discretization of Born-Infeld dynamics, with a squared relativistic speed limit depending on the objective function. This class of frictionless, energy-conserving optimizers proceeds unobstructed until slowing naturally near vanishing loss (up to a self-tunable hyper-parameter shift), which dominates the phase space volume of the system. Building from studies of chaotic systems such as dynamical billiards, we formulate a specific algorithm with good performance on machine learning and PDE-solving tasks, including generalization (so far studied at small scale). In progress are experiments on applications to computational chemistry, sampling, and larger-scale ML, along with further theoretical study of its impact on representation/feature learning. An application of this ML-inspired method to numerical PDE solving for string compactifications blends three of Mike’s many far-reaching insights.

Informations sur la vidéo

  • Date de captation 10/05/2022
  • Date de publication 13/05/2022
  • Institut IHES
  • Langue Anglais
  • Audience Researchers
  • Format MP4

Dernières questions liées sur MathOverflow

Pour poser une question, votre compte Carmin.tv doit être connecté à mathoverflow

Poser une question sur MathOverflow




Inscrivez-vous

  • Mettez des vidéos en favori
  • Ajoutez des vidéos à regarder plus tard &
    conservez votre historique de consultation
  • Commentez avec la communauté
    scientifique
  • Recevez des notifications de mise à jour
    de vos sujets favoris
Donner son avis