TY  - JOUR
AU  - Evans,George W.
AU  - Honkapohja,Seppo
AU  - Williams,Noah
TI  - Generalized Stochastic Gradient Learning
JF  - National Bureau of Economic Research Technical Working Paper Series
VL  - No. 317
PY  - 2005
Y2  - October 2005
UR  - http://www.nber.org/papers/t0317
L1  - http://www.nber.org/papers/t0317.pdf
N1  - Author contact info:
George Evans
Department of Economics
1285 University of Oregon
Eugene, OR 97403-1285
Tel: 541/346-4662
Fax: 541/346-1243
E-Mail: gevans@uoregon.edu

Seppo Honkapohja
Bank of Finland
Finland
E-Mail: Seppo.Honkapohja@bof.fi

Noah M. Williams
Department of Economics
1180 Observatory Drive
University of Wisconsin
Madison, WI 53706-1393
Tel: 608/263-3864
E-Mail: nmwilliams@wisc.edu
AB  - We study the properties of generalized stochastic gradient (GSG) learning in forward-looking models. We examine how the conditions for stability of standard stochastic gradient (SG) learning both differ from and are related to E-stability, which governs stability under least squares learning. SG algorithms are sensitive to units of measurement and we show that there is a transformation of variables for which E-stability governs SG stability. GSG algorithms with constant gain have a deeper justification in terms of parameter drift, robustness and risk sensitivity.
ER  -