-
- Description:
This package performs generalized regression given a set of inputs and outputs. The outputs may represent constraints on the function value or gradient. The constraints can be of inequality or equality. Such learning problems arise when performing classification or regression in the form of inequality constraints and equality constraints respectively, on the function value. This package implements a formulation that combines all such constraints in a single optimization framework.
The formulation follows a SVM methodology by listing down the constraints in a primal problem and computing the corresponding dual which is a convex quadratic program. This program is the solved using the quadprog utility in MATLAB to compute the desired function.
The package also features the $nu$ formulation where it is possible to get a handle on the model complexity in the same way as the classical SVM formulation. A number of examples are included to illustrate these functionalities.
- Changes to previous version:
Initial Announcement on mloss.org.
- BibTeX Entry: Download
- Corresponding Paper BibTeX Entry: Download
- Supported Operating Systems: Windows, Ubuntu
- Data Formats: Any Format Supported By Matlab
- Tags: Svm, Classification, Generalized Regression
- Archive: download here
Comments
No one has posted any comments yet. Perhaps you'd like to be the first?
Leave a comment
You must be logged in to post comments.