Course 2016-2017 a.y.

20231 - BAYESIAN STATISTICAL METHODS


CLMG - M - IM - MM - AFC - CLEFIN-FINANCE - CLELI - ACME - DES-ESS - EMIT - GIO

Department of Decision Sciences

Course taught in English

Go to class group/s: 31
CLMG (6 credits - I sem. - OP  |  SECS-S/01) - M (6 credits - I sem. - OP  |  SECS-S/01) - IM (6 credits - I sem. - OP  |  SECS-S/01) - MM (6 credits - I sem. - OP  |  SECS-S/01) - AFC (6 credits - I sem. - OP  |  SECS-S/01) - CLEFIN-FINANCE (6 credits - I sem. - OP  |  SECS-S/01) - CLELI (6 credits - I sem. - OP  |  SECS-S/01) - ACME (6 credits - I sem. - OP  |  SECS-S/01) - DES-ESS (6 credits - I sem. - OP  |  12 credits SECS-S/01) - EMIT (6 credits - I sem. - OP  |  SECS-S/01) - GIO (6 credits - I sem. - OP  |  SECS-S/01)
Course Director:
PIETRO MULIERE

Classes: 31 (I sem.)
Instructors:
Class 31: PIETRO MULIERE



Course Objectives

The course is intended to be an introduction to Bayesian statistics for students who have already been exposed to a good preliminary course of statistics from a classical point of view.
We also assume that our audience includes those who are interested in using Bayesian methods to model real problems in various scientific disciplines (economics, finance, econometrics, survival analysis, demography, reliability, etc.)
Such students usually want to understand foundational principles well enough so that they feel comfortable using statistical procedures; are able to recommend solutions based upon these procedures to decision makers; are intrigued enough to seek additional background understanding through relevant literature.
For this reason, we have tried to maximize interpretation of the theory and have minimized our dependence upon proof of theorems.


Course Content Summary

  • Subjective probability:
    • Definition of coherence; definition of probability; existence of a probability; properties.
  • The Bayesian Methods:
    • Bayes' Theorem; Posterior Distributions; Prior Distributions; Decision Theory.
  • Inference:
    • Predictive and parametric inference; Coniugate Analysis; Parametric point estimation; Test of Hypotheses.
  • Exchangeable sequences of random variables:
    • Exchangeability events; mixtures of i.i.d. random variables and exchangeability; De Finetti's representation theorem; partial exchangeability according to De Finetti; partial exchangeability according to Diaconis and Freedman.
  • Predictive sufficient statistics:
    • Predictive sufficient statistic: definitions and properties; connections between predictive sufficient statistics and statistical models; predictive sufficient statistics and exchangeability; exchangeability and characterizations of the models; predictive justifications of the priors.
  • Prior processes for the Bayesian non-parametric:
    • Constructing random probability measures; the Dirichlet process; Polya Trees; neutral to the right priors.
  • Urn schemes for constructing priors:
    • Background on two colors urn processes; a Polya urn for the Dirichlet process; reinforced urn processes; reinforced urn processes for the beta-Stacy prior. Reinforced urn processes for the Polya tree; neutral to the prior.
  • Linear models; Kalman filter.

Detailed Description of Assessment Methods

The exam consists of an essay on a topic chosen by the student, which is discussed with the professor.


Textbooks

Teaching material is available in yoU@B Diary.

Exam textbooks & Online Articles (check availability at the Library)
Last change 21/03/2016 12:31