Standard Article

Nonparametric Curve Estimator

Statistical and Numerical Computing

  1. Michael E. Tarter

Published Online: 15 JAN 2013

DOI: 10.1002/9780470057339.vnn129

Encyclopedia of Environmetrics

Encyclopedia of Environmetrics

How to Cite

Tarter, M. E. 2013. Nonparametric Curve Estimator. Encyclopedia of Environmetrics. 4.

Author Information

  1. University of California, Berkeley, CA, USA

Publication History

  1. Published Online: 15 JAN 2013

Abstract

The roles that nonparametric curve estimators can play are illustrated in studies of polluting chemicals that have built up in the environment and of crop forecasting techniques. The multivariate normal model and model-free curve estimator representations are compared. Equations and approaches are described that enable researchers to input prior information that divides variates into two categories, variates of primary importance and variates that are nuisances in the sense that their detained distributional structure is relevant solely insofar as its effects on variates of primary importance. It is shown that conditional density estimators can be computed given user-selected values of nuisance variates, not computed as a second step that follows a first step, which requires the estimation of a high-dimensional joint distribution. It is argued that this approach resembles parametric methodology that has long been fruitfully employed in the field of experimental design.

Keywords:

  • curse of dimensionality;
  • data transformation;
  • edge-effects;
  • Fourier series;
  • kernel;
  • kitchen ventilation;
  • indoor air pollution;
  • nursing infant milk intake;
  • orthogonal series;
  • vineyard