Estimating Stochastic Differential Equations Using Repeated Eigenfunction Estimation and Neural Networks

We propose identifying the drift and the diffusion functions of an ergodic scalar stochastic differential equation using repeated eigenfunction estimation. The transition density will be estimated in a new way involving Kolmogorov’s backward equation, neural networks and functions of our choice. Martingale estimating functions will be used to obtain asymptotic properties.

MoreLess

Year of publication:	2012-10-24
Authors:	Tuncer, Ruhi
Institutions:	Galatasaray Üniversitesi İktisadi Araştırmalar Merkezi (GİAM), Galatasaray Üniversitesi
Subject:	Stochastic Differential Equation \| Kolmogorov’s backward Equation \| Infinitesimal Generator \| Eigenfunctions \| Transition Density \| Neural Networks \| Martingale Estimating Functions

More details

Extent:	application/pdf
Series:	GIAM Working Papers.
Type of publication:	Book / Working Paper
Notes:	The text is part of a series GIAM Working Papers Number 12-5 15 pages
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10010840310