Projet de statistique bayésienne sur l'article :
Kaufmann, S. (2010). Dating and forecasting turning points by Bayesian clustering with dynamic structure: a suggestion with an application to Austrian data, Journal of Applied Econometrics, 25, 309 – 344.
Disponible ici : https://arkensae.github.io/BayesianTurningPoint/Biblio/Kaufmann%2C%20S.%20(2010).pdf
Date de rendu : 30 janvier 2021 8h.
Consignes :
-
The project length has to be about 5 pages (plus code), with reasonable font size and margins.
-
It is an applied project. Each group has to choose a statistical application with real or simulated data and then analyse it with Bayesian methods related to the material seen in class (or extensions of it).
-
A possible structure of the project is: (1) definition of the problem under study and explanation of why it is interesting, (2) choice of the appropriate Bayesian technique (you should explain the methodology used and the motivation why you have chosen it), (3) de- scription of the computational method used and difficulties encountered, (3) explanation and interpretation of the results. You can also compare the results you find with the results that you find with a frequentist approach.
-
If you choose to develop a project based on an academic article you can, for instance, either replicate and extend the simulations in the paper, or find a real data set and apply the method of the paper to it, or develop a simulation study if the paper does not contain it. An article can be chosen only by one group on the first-in-first-out basis. So, if you want to choose an article among the ones proposed below please email me the titles of three articles in order of preference and I will assign you the first article in your list that is still available.