From Latent Dynamics to Meaningful Representations.
J Chem Theory Comput
; 20(9): 3503-3513, 2024 May 14.
Article
de En
| MEDLINE
| ID: mdl-38649368
ABSTRACT
While representation learning has been central to the rise of machine learning and artificial intelligence, a key problem remains in making the learned representations meaningful. For this, the typical approach is to regularize the learned representation through prior probability distributions. However, such priors are usually unavailable or are ad hoc. To deal with this, recent efforts have shifted toward leveraging the insights from physical principles to guide the learning process. In this spirit, we propose a purely dynamics-constrained representation learning framework. Instead of relying on predefined probabilities, we restrict the latent representation to follow overdamped Langevin dynamics with a learnable transition densityâa prior driven by statistical mechanics. We show that this is a more natural constraint for representation learning in stochastic dynamical systems, with the crucial ability to uniquely identify the ground truth representation. We validate our framework for different systems including a real-world fluorescent DNA movie data set. We show that our algorithm can uniquely identify orthogonal, isometric, and meaningful latent representations.
Texte intégral:
1
Collection:
01-internacional
Base de données:
MEDLINE
Langue:
En
Journal:
J Chem Theory Comput
/
J. chem. theory comput. (Online)
/
Journal of chemical theory and computation (Online)
Année:
2024
Type de document:
Article
Pays d'affiliation:
États-Unis d'Amérique
Pays de publication:
États-Unis d'Amérique