Premium
GENERALIZED STOCHASTIC GRADIENT LEARNING *
Author(s) -
Evans George W.,
Honkapohja Seppo,
Williams Noah
Publication year - 2010
Publication title -
international economic review
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.658
H-Index - 86
eISSN - 1468-2354
pISSN - 0020-6598
DOI - 10.1111/j.1468-2354.2009.00578.x
Subject(s) - robustness (evolution) , computer science , convergence (economics) , stability (learning theory) , rational expectations , stochastic approximation , mathematics , mathematical optimization , artificial intelligence , econometrics , machine learning , economics , key (lock) , biochemistry , chemistry , computer security , gene , economic growth
We study the properties of the generalized stochastic gradient (GSG) learning in forward‐looking models. GSG algorithms are a natural and convenient way to model learning when agents allow for parameter drift or robustness to parameter uncertainty in their beliefs. The conditions for convergence of GSG learning to a rational expectations equilibrium are distinct from but related to the well‐known stability conditions for least squares learning.