The regression methods attempt to fit a line to the historical data. The following definitions provide the mathematical equations for defining the lines.
For all equations, y is the forecast value and x is the historical time series.
Linear Regression Method
A linear relationship (y=a*x+b) is fitted to the data.
Where a is the slope and b is the intercept.
Nonlinear Regression Methods
In the nonlinear regression methods, a linear relationship (y'=a*x'+b) is fitted to a transformation of the original data. Each method uses a different equation for the transformation.
Polynomial Fit:The equation x'=log(x) and y'=log(y) results in the development of a polynomial model between x and y (y=c*x^a).
Exponential Fit: The equation x'=x and y'=ln(y) results in the development of an exponential model between x and y (y=c*e^ax).
Logarithmic Fit: The equation x'=log(x) and y'=y results in the development of a logarithmic model between x and y (y=a*log(x)+b).
Asymptotic Fit:The equation x'=1/x and y'=1/y results in the development of an asymptotic curve (y=x/(a+bx)).
Exponential Asymptotic Fit: The equation x'=x and y'=ln(y/(K-y)) results in the development of an exponential asymptotic curve (y=cKe^ax/(1+ce^ax)) .
Copyright © 2003, 2007, Oracle. All rights reserved.