Bayesian Methods in Reliability na UFMG

[loschi@est.ufmg.br]

O professor Fabrizio Ruggeri do Instituto di Matemática Applicata e
Tecnologie Informatiche (IMATI) de Milão, Italia, estara visitando o 
Departamento de Estatistica da UFMG no periodo de 03 a 06 de abril de 2006
logo apos o 8o. EBEB). Neste periodo, o professor Fabrizio ministrara o 
minicurso titulado "Bayesian Methods in Reliability" cujo programa encontra-se 
a seguir.

Alem de Research Director do IMATI, o professor Fabrizio e´ presidente do ENBIS 
(European Network for Business and Industrial Statistics), e´ membro ativo da  
International Society for Bayesian Analysis (ISBA) e  tem contribuido muito 
para difundir o uso de tecnicas bayesianas nas industrias  europeias.

O horario do minicurso e´ de 9h `as 12h, na sala 2076 do ICEx. Inscricoes podem 
ser feitas na Secretaria do Departamento de Estatistica com Cristina 
(cristina@est.ufmg.br) e Rose (rosiane@icex.ufmg.br). 

Maiores informacoes podem ser obtidas pelos telefones (31) 3499 5900 e 
(31) 3499 5920.

Um abraco a todos
Rosangela 

* * * 

BAYESIAN METHODS IN RELIABILITY 

The key aspects of the statistical analysis of reliability will be illustrated 
from basics to some applications (e.g. gas escapes, train failures). A Bayesian 
approach will be followed in the statistical inference, although some result 
from the frequentist viewpoint will be illustrated. Both parametric and 
nonparametric methods will be used".

Outline: 
1) Reliability: history, different scenarios, link with survival analysis.

2) Basic definitions and results (e.g. survival functions, hazard rate, mean 
time to failure, Poisson processes and their properties).

3) Nonrepairable systems: definition, examples, statistical analysis.

4) Repairable systems: definition, examples, renewal and Poisson processes.

5) Homogeneous Poisson processes: inference and gas escape example.

6-7) Nonhomogeneous Poisson processes: inference and gas escape and train 
failure example (parametric and nonparametric approaches).

8-9) Accelerated Failure Time: parametric and nonparametric approaches