Fit Automatically Selected by Training

Selects the best available Fit from a list of available methods you have chosen. In addition to selecting the best method, Fit Automatically Selected by Training also automatically adjusts the individual settings (often called hyperparameters) to find the optimizing, predictive performance while avoiding overfitting.

Usability Characteristics

  • Fits both noisy and non-noisy data.
  • Reduces the methods on which Fit Automatically Selected by Training iterates in order to reduce the run time used to build the Fit.
  • Can run in multi-execute, while simultaneously iterating over multiple responses.
  • The Stepwise Regression Terms option for Least Squares Regression reduces the number of coefficients in the regression model to contain only the set that is statistically significant.
  • The behavior and characteristics of the underlying methods are the same as when the methods are directly applied. See their respective documentation pages for details.
  • Gradient information can be used to boost performance for the methods that support gradients.

Settings

In the Specifications step, Settings tab, change method settings.
Parameter Default Range Description
Least Square Regression On On or Off
On
Use Stepwise Regression Terms to reduce the number of terms in the regression to the statistically significant set.
Off
Do not consider Least Squares Regression in the ensemble list of methods.
Stepwise Regression Terms Full Quadratic
  • Linear
  • Squared
  • Cubic
  • Interaction
  • Full Quadratic
  • Full Cubic
Controls the maximal set of terms considered in stepwise Least Squares Regression.
Linear
First order terms only.
y=A+Bx+Cy
Squared
Second order without cross terms.
y=A+Bx+Cy+Dx^2+Ey^2
Cubic
Third order without cross terms.
y=A+Bx+Cy+Dx^2+Ey^2+Fx^3+Gy^3
Interaction
Linear and the cross terms.
y=A+Bx+Cy+Dxy
Full Quadratic
Complete second order polynomial.
Full Cubic
Complete third order polynomial.
Moving Least Squares On On or Off
On
Consider Moving Least Squares Method in the ensemble list of methods.
Off
Do not consider Moving Least Squares Method in the ensemble list of methods.
Radial Basis Function On On or Off
On
Consider Radial Basis Function in the ensemble list of methods.
Off
Do not consider Radial Basis Function in the ensemble list of methods.
Use Gradient Data On On or Off
On
Allow methods to be enhanced by gradient information when it is available.
Off
Do not allow methods to be enhanced by gradient information.