A mex interface for SMO style SVM+ solvers developed on top of LIBSVM package In [1] and [2], the code follows the same structure of the original LIBSVM mex interface. The original code developed by Dmitry Pechyony is also provided in the orignal code folder.
Before running the make.m script, check for a valid C/C++ compiler installed in your system; once you have passed this step just execute the make.m script and it will build the mex functions.
- 64-bits windows compiled mex functions are provided in this repo, Microsoft Visual C++ 2012 was used as a compiler (Microsoft Visual C++ 2013 was tested succussfully).
- The code was tested on Matlab R2014a and R2015a.
Run the svm_plus_demo.m script
Please refer to the issues section
Coming soon
[1] D. Pechyony, R. Izmailov, A. Vashist and V. Vapnik. SMO-style Algorithms for Learning using Privileged Information . DMIN 2010.
[2] D. Pechyony and V. Vapnik. Fast Optimization Algorithms for Solving SVM+. Chapter in Statistical Learning and Data Science, 2011.