VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.

401-4624-00L 5 Credits

Selected chapters in time series analysis

AK Zeitreihenanalyse

VVZ CR n/a

Last Updated: 2026-02-05 15:02:32

Abstract

Introduction to state space models. Filtering, smoothing and parameter estimation: Exact recursive computations in discrete and linearGaussian models. Markov chain Monte Carlo and particle filter methods for general state space models.Introduction to long range dependece.

Content

The main topic of this course is the use of state space models in time series analysis. These models assume that the observations are incomplete and noisy functions of an unobservable state process with a simple Markovian dynamics. A first part will illustrate the flexibilty of this approach by a number of examples. Then I deal with methods to compute the likelihood and conditional distributions of the state given a stretch of observed values. For linear Gaussian models and for models with discrete state space, recursions can be computed in closed form whereas for more general models one needs some approximations. I will concentrate on sequential Monte Carlo approximations that have received much attention recently. Depending on time and the interest of students, I might discuss models with long range dependence towards the end of the course.

Resources

Literature

J. Durbin and S. J. Koopman, Time Series Analysis by State Space Methods. Oxford University Press, 2001. H. R. Künsch, State Space and Hidden Markov Models. In Complex Stochastic Systems, O.E. Barndorff-Nielsen, D. Cox and C. Klüppelberg, eds. Chapman and Hall,/CRC 2001, 109-173.

General Information

Language
English

Examination

Type
session examination
Mode
oral 30 minutes

Course Components

Type Title Time & Place Hours
lecture AK Zeitreihenanalyse
  • Wed 10:15-12:00 (HG D 1.1)
2 h weekly

Offered In