Ads by Google
Christian Borgelt's Web Pages

MPR - Multivariate Polynomial Regression


32 bit 64 bit (32/64 bit only for executables) (386 kb) (591 kb) GNU/Linux executables (203 kb) (228 kb) Windows console executables (189 kb) regress.tar.gz (161 kb) C sources, package version 4.4 (2015.11.30)


MPR is a program for multivariate polynomial regression, i.e., a program that determines a minimum squared error polynomial for given data. The output values may also be transformed with the logit transformation, thus enabling (multivariate) logistic regression. Alternatively, a logistic regression may be obtained with an iterative scheme that maximizes data likelihood. Since version 4.0 it also allows for a forward selection of attributes based on a separate validation data set. A brief description of how to apply this program can be found in the file regress/ex/readme in the source package.

If you have trouble executing the programs on Microsoft Windows, check whether you have the Microsoft Visual C++ Redistributable for Visual Studio 2017 (see under "Other Tools and Frameworks") installed, as the programs were compiled with Microsoft Visual Studio 2017.

A graphical user interface to these command line programs (RegGUI, written in Java) can be found here. This graphical user interface contains a simple viewer for induced regression models, which is also available as a separate program. This graphical user interface does not yet support logistic regression based on maximum likelihood or attribute selection with a validation data set.