Title :
Quantization for Nonparametric Regression
Author :
Györfi, László ; Wegkamp, Marten
Author_Institution :
Budapest Univ. of Technol. & Econ., Budapest
Abstract :
The authors discuss quantization or clustering of nonparametric regression estimates. The main tools developed are oracle inequalities for the rate of convergence of constrained least squares estimates. These inequalities yield fast rates for both nonparametric (unconstrained) least squares regression and clustering of partition regression estimates and plug-in empirical quantizers. The bounds on the rate of convergence generalize known results for bounded errors to subGaussian, too.
Keywords :
convergence; least mean squares methods; nonparametric statistics; quantisation (signal); regression analysis; bounded error; convergence; mean squared error method; nonparametric least squares regression; vector quantization; Additive noise; Convergence; Data compression; Least squares approximation; Multivariate regression; Probability distribution; Temperature; Vector quantization; Weather forecasting; Yield estimation; Regression estimation with restriction; finite-sample bounds; least squares estimates; vector quantization;
Journal_Title :
Information Theory, IEEE Transactions on
DOI :
10.1109/TIT.2007.913565
