Statistical properties of nearest neighbor regression function estimate for strong mixing processes.

Patrick Rakotomarolahy (1)

1 CES - Centre d'économie de la Sorbonne

We examine statistical properties of nearest neighbor regression technique beyond the standard i.i.d hypothesis. We analyze the second order properties and establish the asymptotic normality of the nearest neighbor regression function estimate for strongly mixing processes. We achieve some rates for the second order properties for this nonparametric method for strongly mixing processes. We make an application of the theoretical results on the modelling of economic indicators.

Mots clés: nearest neighbor; strong mixing processes; second order properties; asymptotic normality; economic indicators

Domaines: Applications [stat.AP]

Patrick Rakotomarolahy. Statistical properties of nearest neighbor regression function estimate for strong mixing processes.. 2012. ⟨hal-00725841⟩