WellSVM is a package for scalable and convex weakly labeled SVMs. The package includes the MATLAB code of the algorithm WellSVM.
Yu-Feng Li, Ivor Tsang, James Kwok and Zhi-Hua Zhou. Scalable and convex weakly labeled SVMs
. Journal of Machine Learning Research, 14:2151-2188, 2013.
This package is free for academic usage. You can run it at your own risk. For other purposes, please contact Prof. Zhi-Hua Zhou
The package was developed with MATLAB(v2008a) and C++, hence Matlab and C++ need to be installed before running this code. A demo function and some data sets are included to illustrate the use of this package.
This package was developed by Mr.Yu-Feng Li
(email@example.com). For any problem concerning the code, please feel free to contact Mr. Li.
: code & data
(2.7 MB), for semi-supervised learning, regular scale data setscode & data
(40 MB), for semi-supervised learning, large scale data setscode & data
(814 KB), for clustering, regular scale data setscode
(141 KB) & data
(456 MB), for clustering, large scale data setscode for multiple instance learning