Abstract
This article focuses on estimating an autoregressive regression model for circular time series data. Simulation studies have shown the difficulties involved in obtaining good estimates from low concentration data or from small samples. It presents an application using real data.
Mathematics Subject Classification:
Acknowledgment
The second author acknowledges the support of FAPESP (grants 03/10105-2 and 08/51097-6).
Notes
V(·): absolute bias estimate (in percentege); MSE (·): mean square error estimate; Convergence: number of cases in which the algorithm converged.
rims.unh.edu accessed in January 2006.