CONVERGENCE OF REPETITIVE GUESSTIMATION ALGORITHM ON A LINEAR REGRESSION PROBLEM

Repetitive Stochastic Guesstimation (RSG) is a probabilistic algorithm introduced in Charemza (1996) which mimics the usual guessing of the parameters involved in a complex large macroeconomic model. The paper analyzes the objective function employed by RSG and it establishes sufficient conditions on the RSG components in order to achieve global convergence for a specific task, namely the linear regression problem.

MoreLess

Year of publication:	2001
Authors:	Agapie, Adriana
Published in:	Journal for Economic Forecasting. - Institutul de Prognoza Economica. - 2001, 2, p. 118-126
Publisher:	Institutul de Prognoza Economica
Subject:	estimation \| macromodels \| computational techniques \| methodology

More details

Type of publication:	Article
Classification:	C6 - Mathematical Methods and Programming ; C63 - Computational Techniques
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005612235