We consider in this article a class of nonlinear additive autoregressi
ve models with exogenous variables for nonlinear time series analysis
and propose two modeling procedures for building such models. The proc
edures proposed use two backfitting techniques (the ACE and BRUTO algo
rithms) to identify the nonlinear functions involved and use the metho
ds of best subset regression and variable selection in regression anal
ysis to determine the final model. Simulated and real examples are use
d to illustrate the analysis.