de.mpg.escidoc.pubman.appbase.FacesBean

Post

 
 Vis
  A Comparison of Frequentist and Bayesian Approaches to Econometric Modelling
Item is

Ophav

 Ophav:
Lohmeyer, Jan Hendrik1, Forfatter
Bohn Nielsen, Heino2, Vejleder
Tilknytninger:
1Det Samfundsvidenskabelige Fakultet, Københavns Universitet, København, Danmark, diskurs:7001              
2Økonomisk Institut, Det Samfundsvidenskabelige Fakultet, Københavns Universitet, København, Danmark, diskurs:7014              
skjul Ophav
Vis Ophav

Indhold

Ukontrollerede emneord: Bayesian, Frequentist, Smooth Particle Filter, MCMC, Stochastic Volatility, Metropolis-Hastings
 Abstract: Bayesian and frequentist methods are complementing econometric approaches each offering a wide range of different tools that the econometrician can choose from. Particle filters are represented in both categories. They can be thought of as generalisations of the Kalman filter.

Inspired by two recent advances on the application of particle filters for stochastic volatility (SV) models I study state-space models (SSM) in my thesis: Flury and Shephard (2011) introduced the Bayesian particle Markov chain Monte Carlo (PMCMC) method into the realm of SV estimation methods. Malik and Pitt (2009) showed that the resampling step in a particle filter can be designed in a smooth manner, such that frequentist simulated maximum-likelihood (SML) estimation is feasible. The applied part of this thesis first covers a simplified (linear Gaussian) state-space model to prepare the comparison and discussion of both approaches applied to financial data.
skjul Indhold
Vis Indhold

Filer

Navn:
Lohmeyer.pdf (Hovedtekst)
Bemærkninger:
-
Tilgængelighed:
Offentlig
Mime-type / størrelse:
application/pdf / 982KB
Copyright dato:
2013-10-25
Copyright information:
De fulde rettigheder til dette materiale tilhører forfatteren.
skjul Filer
Vis Filer

Basal

Bogmærk denne post: https://diskurs.kb.dk/item/diskurs:57024:3
 Type: Speciale
Alternativ titel: The Case of Particle Filters
skjul Basal
Vis Basal

Links

Vis Links

Detaljer

Sprog: English - eng
 Datoer: 2013-06-17
 Sider: -
 Publiceringsinfo: København : Københavns Universitet
 Indholdsfortegnelse: -
 Note: -
 Type: Speciale
skjul Detaljer
Vis Detaljer

Kilde

Vis Kilde