Kernel regression estimation with np PDF

Preview

Summary

Download Kernel regression estimation with np PDF

Description

Kernel regression estimation with np The np package provides a complete framework for performing a more sophisticated nonparametric regression estimation for local constant and linear estimators, and for computing cross-validation bandwidths. The two workhorse functions for these tasks are np::npreg and np::npregbw, which they illustrate the philosophy behind the np package: first, a suitable bandwidth for the nonparametric method is found (via np::npregbw) and stored in an rbandwidth object; then, a fitting function ( np::npreg) is called on that rbandwidth object, which contains all the needed information. # Data -- nonlinear trend data(Auto, package = "ISLR") X...