Fitting a kernel curve to your data is appropriate for situations where little is known about the process under study. To fit such a curve, you specify the smoothness of the fit, and then the data actually determine the functional fit of the curve.
The smoothing parameter takes values greater than 0 and less than or equal to 100. The higher the value of the smoothing parameter, the smoother the curve. A very high value results in a curve that is close to a straight line, while smaller values produce lines with more curvature. By default, the MISE is used as the smoothing parameter.
Three types of kernel functions can be used: normal, triangular, and quadratic. The normal function is the default.