A nonparametric goodness-of-fit test for a class of parametric autoregressive models

✍ Scribed by Joseph Ngatchou Wandji

• Publisher: Elsevier Science
• Year: 1998
• Tongue: English
• Weight: 172 KB
• Volume: 71
• Category: Article
• ISSN: 0378-3758
• DOI: 10.1016/s0378-3758(98)00010-x

✦ Synopsis

We derive a nonparametric test for discriminating between generalized autoregressive models. This test is a modiÿcation of the nonparametric tests proposed in Diebolt and Ngatchou Wandji (Prà epublications Mathà ematiques de l'Università e Paris-Nord, vol. 96-04) and Diebolt et al. (Scand. J. Statist. 24,[241][242][243][244][245][246][247][248][249][250][251][252][253][254][255][256][257][258][259]. It is based on a suitably normalized sum of residuals. The null distribution and the power of the test under both ÿxed and a sequence of local alternatives are studied under mild stationarity and -mixing conditions. This procedure can be applied to testing linear models against nonlinear models or certain nonlinear models against others. Numerical simulations show that the proposed test is powerful against most of the alternatives considered.