Generalized normal ARMA model applied to the areas of economy, hydrology, and public policy

ABSTRACT

The main purpose of this article is the presentation of a new class of time series models which is the merge output of the generalized normal distribution with ideas from the GARMA model. Symmetrically, tails that may be lighter or heavier than the Gaussian distribution, and Gaussian and Laplace distributions as special cases, are the main advantages of the use of generalized normal distribution. The proposed model is called generalized normal autoregressive moving average (GN-ARMA). We exemplify the application of the proposed model adjusting it to the three time series, which are from the areas of economy, hydrology, and public policy.

KEYWORDS:

• ARMA models
• Generalized normal distribution
• Time series

MATHEMATICS SUBJECT CLASSIFICATION:

• 37M10
• 62M10

Acknowledgment

The studies of the first author was funded by Capes-Brazil.