Advances and Applications in Statistics
Volume 20, Issue 1, Pages 1 - 23
(January 2011)
|
|
TWO MULTIVARIATE STOCHASTIC VOLATILITY MODELS APPLIED TO AIR POLLUTION DATA FROM S�O PAULO, BRAZIL
Jorge A. Achcar, Henrique C. Zozolotto, Eliane R. Rodrigues and Paulo H. Nascimento Saldiva
|
Abstract: In this paper, we use some recently introduced multivariate stochastic volatility models to problems related to air pollution. The models considered here are commonly used in studies of financial time series. In this paper, they are used to analyse the weekly average volatility of five pollutants affecting the inhabitants of the city of S�o Paulo, Brazil. Two different models are proposed to explain the behaviour of the weekly average measurements of those pollutants. Those models depend upon some parameters that are estimated using a Bayesian formulation via Markov chain Monte Carlo (MCMC) methods. |
Keywords and phrases: Bayesian inference, MCMC methods, oxides, ozone, particulate matter. |
|
Number of Downloads: 372 | Number of Views: 1129 |
|