Fast indirect robust generalized method of moments

Authors:
Sébastien Loisel;Marina Takane
Affiliations:
Department of Mathematics, 1805 North Broad Street, Temple University, Philadelphia, PA 19122, United States;1003 Christian Street #2, Philadelphia, PA 19147, United States
Venue:
Computational Statistics & Data Analysis
Year:
2009

Citing 2
Cited 2

Extrapolation methods for vector sequences

SIAM Review
High strong order explicit Runge-Kutta methods for stochastic ordinary differential equations

Applied Numerical Mathematics - Special issue celebrating the centenary of Runge-Kutta methods

Fast robust estimation of prediction error based on resampling

Computational Statistics & Data Analysis
Generalized GIPSCAL re-revisited: a fast convergent algorithm with acceleration by the minimal polynomial extrapolation

Advances in Data Analysis and Classification

Quantified Score

Hi-index	0.03

Visualization

Abstract

The Robust Generalized Methods of Moments (RGMM) and the Indirect Robust GMM (IRGMM) are algorithms for estimating parameter values in statistical models, such as diffusion models for interest rates, in a robust way. The long computation time is one of the main challenges facing these methods. In this paper, we introduce accelerated variants of RGMM and IRGMM. The fixed point iteration in RGMM is accelerated using minimal polynomial extrapolation, and the simulation of pseudo-observations in IRGMM is sped up by using a higher order stochastic Runge-Kutta method. We illustrate the fast performance of these algorithms for an interest rate diffusion model on four datasets.