## Applied Statistics for EORAS

 Faculteit Economie en Bedrijfskunde Jaar 2017/18 Vakcode EBM805B05 Vaknaam Applied Statistics for EORAS Niveau(s) master Voertaal Engels Periode semester II a ECTS 5 Rooster rooster

Uitgebreide vaknaam Applied Statistics for EORAS
Leerdoelen Upon completion of the course the student is able to:
1. Appraise the notion of subjective probability and its consequences for inference via Bayes’ theorem.
2. Devise ways to implement advanced Bayesian methods on a computer.
3. Conduct inference using simulation methods such as Gibbs sampling or the Metropolis-Hastings algorithm.
4. Evaluate the performance of such methods by means of e.g. Monte Carlo methods.
5. Formulate a research hypothesis and design an appropriate Bayesian econometric approach to empirically investigate the hypothesis.
6. Alternatively or additionally, evaluate the robustness of findings on research hypotheses in recent publications by assessing the hypotheses using Bayesian empirical methods.
7. Present the approach in a term paper and an oral presentation.
Omschrijving We give an overview of the Bayesian approach to statistics and econometrics. Unlike “conventional” statisticians and econometricians, Bayesians are subjective need not believe in the notion of a fixed true parameter, but instead formulate their beliefs on parameters of interest and use data to update these beliefs. The workhorse tool to achieve this is Bayes' theorem. We shall devote attention to both the theoretical underpinnings of the methods and their practical implementation to (dynamic) econometric models.
We start with basics such as Bayes’ theorem, prior and posterior distributions, then discuss the workhorse linear regression model from a Bayesian point of view to move on to simulation methods like rejection sampling, Gibbs sampling and the Metropolis-Hastings algorithm. The final part applies these methods to some (dynamic) econometric models.
Uren per week 6
Onderwijsvorm opdracht(en), werkcolleges, zelfstudie
Toetsvorm opdracht(en), presentatie(s)
(Individual assignments will be: 1. exercises that deepen students' knowledge of the methods discussed in the course and 2. a term paper studying or applying the methods of the course.)
Vaksoort master
Coördinator prof. dr. R.H. Koning
Docent(en) dr. C.H. Hanck
Verplichte literatuur
Titel Auteur ISBN Prijs
(Additional literature) Bayesian Computation with R. Use R!, 2nd edition, 2010, New York: Springer Albert, J.
(Additional literature) Bayesian Econometric Methods. Econometrics Exercises, 2007, Cambridge: Cambridge University Press Koop, G., D.J. Poirier, J.L. Tobias (eds)
(Additional literature) Bayesian Econometrics, 2003, New York: Wiley Koop, G.
(Additional literature) Bayesian Inference in Dynamic Econometric Models. Advanced Texts in Econometrics, 2000, Oxford: Oxford University Press Bauwens L., M. Lubrano, J.F. Richard
(Additional literature) Markov chain Monte Carlo methods: Computation and inference, 2001, in Heckman J.J., E.E. Leamer (eds.) Handbook of Econometrics, Amsterdam: Elsevier v.5, ch.57 Chib, S.
(Additional literature) The Oxford Handbook of Bayesian Econometrics, 2011, Oxford: Oxford University Press various authors
(Compulsory literature) Introduction to Bayesian Econometrics, 2008, Cambridge: Cambridge University Press Greenberg, E. 9781107436770 €  30,59
Entreevoorwaarden Statistics: Statistical Inference, Casella, G.C, R.L. Berger
Econometrics: Econometrics, Hayashi, F., ch. 1-3
Opmerkingen Secretariaat: Martine Geerlings-Koolman, tel.: 050 3637018, email: m.a.koolman@rug.nl.
Opgenomen in
Opleiding Jaar Periode Type
MSc Econometrics, Operations Research & Actuarial Studies/EORAS  (keuzevakken MSc EORAS) 1 semester II a keuze