PORTALE DELLA DIDATTICA

Ricerca CERCA
  KEYWORD

Area Ingegneria

Knowledge-informed Machine Learning

Riferimenti DANIELE APILETTI

Descrizione The goal is to understand, manipulate, and improve Physics-informed neural networks (PINNs). Through deep learning, these algorithms allow the solution of differential equations.
PINNs are well-known to suffer from many limitations when facing complex systems, then a prominent direction in research is to develop new ways of training them.
The candidate will apply these innovative algorithms to various well-known hard problems to develop new generalized learning strategies.


Scadenza validita proposta 19/02/2025      PROPONI LA TUA CANDIDATURA