FindGraph provides an easy way to determine the bestfit parameters for
linear regression model.
The form of the general leastsquares linear regression model is:
where f_{j}(X) are any arbitrary functions of X.
In regression modeling, the term 'linear'
means that the models dependence on its parameters Aj is linear. The functions
f_{j}(X) may be nonlinear.
In FindGraph, linear regression model is linear combination of
Polynomial, Rational, Logarithmic, Exponential, and Fourier functions f_{jk}(X).
The parameters A_{j} are estimated by the method of leastsquares
to minimize the difference between the model and data.
The Wizard of Approximation will help you to find the best equation
and get a report of the results in seconds.
There is feature to apply robust fitting instead linear regression.
After regression we exclude points out of 1.5*stdErr interval, and find regression line again.
There is feature to apply regularization, mainly Forward Stagewise Linear Regression algorithm to prevent overfitting.
FindGraph provides automatic logging of all curvefit analysis.
Read more about:
Digitizing
Graphing
Curve fitting
Bestfit equation
Closed curves
Empirical models
Library
Analysis
Tools
Filters
Convolution
Periodic signals
Procrustes analysis
Multipeak fitting
SSA forecasting
