Virtually every organism gathers information about its noisy environment and builds models from those data, mostly using neural networks. Here, we use stochastic thermodynamics to analyze the learning of a classification rule by a neural network. We show that the information acquired by the network is bounded by the thermodynamic cost of learning and introduce a learning efficiency η≤1. We discuss the conditions for optimal learning and analyze Hebbian learning in the thermodynamic limit.

Stochastic Thermodynamics of Learning / Goldt, S.; Seifert, U.. - In: PHYSICAL REVIEW LETTERS. - ISSN 0031-9007. - 118:(2017), pp. 1-5. [10.1103/PhysRevLett.118.010601]

Stochastic Thermodynamics of Learning

Goldt S.;
2017-01-01

Abstract

Virtually every organism gathers information about its noisy environment and builds models from those data, mostly using neural networks. Here, we use stochastic thermodynamics to analyze the learning of a classification rule by a neural network. We show that the information acquired by the network is bounded by the thermodynamic cost of learning and introduce a learning efficiency η≤1. We discuss the conditions for optimal learning and analyze Hebbian learning in the thermodynamic limit.
2017
118
1
5
010601
Goldt, S.; Seifert, U.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11767/117827
Citazioni
  • ???jsp.display-item.citation.pmc??? 3
  • Scopus 36
  • ???jsp.display-item.citation.isi??? 32
social impact