An introduction to the main Monte Carlo methods for Bayesian inference, such as MC integration, sampling-importance-resampling (SIR), Markov chain Monte Carlo (MCMC) and sequential MC (SMC), will also be introduced. The inferential approach of this course is predominantly Bayesian, so we will briefly introduce key ingredients of Bayesian inference, model selection and criticism. Multivariate time series models we will considere include vector autoregressive (VAR) models, factor-augmented VARs, dynamic factor models and various time-varying covariance models. ![]() Univariate time series models we will consider include the family of autoregressive (fractionally) integrated moving average (ARIMA) models, dynamic linear models (aka state-space) models, Markov switching models, generalized autoregressive conditionally heteroskedastic (GARCH) and stochastic volatility (SV) models. A strong background in calculus, probability, statistics and matrix algebra is highly beneficial.Ĭourse description: The main goal of the course is to make the student familiar with and able to implement univariate and multivariate modern time series models. Office hours: Wednesdays, from 11am to 12pm (by appointment only)ĪTTENTION: This is an advanced time series course (see course description below!). ![]() ![]() Professor: Hedibert Freitas Lopes - Lectures: Mondays and Wednesdays, from 10:30am to 11:45am (January 10th to April 27th)
0 Comments
Leave a Reply. |