|
George W. Evans, Seppo Honkapohja, Noah Williams
NBER Technical Working Paper No. 317
Issued in October 2005
NBER Program(s): TWP
---- Abstract -----
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.
Would you like an annual subscription to NBER Working Papers? Click
here for more information.
You may purchase this paper on-line in .pdf format
from SSRN.com ($5) for electronic delivery.
Information for subscribers and others expecting no-cost downloads
Machine-readable bibliographic record -
MARC,
RIS,
BibTeX
| 
