S4VM算法(Matlab)

------------------------------------------------------------------------------------------ Readme for the S4VM Package version May 9, 2011 ------------------------------------------------------------------------------------------ The package includes the MATLAB code of the semi-supervised algorithm S4VM, which towards making unlabeled data never hurt, or safe semi-supervised algorithm [1]. [1] Y.-F. Li, Z.-H. Zhou. Towards Making Unlabeled Data Never Hurt. In: Proceedings of the 28th International Conference on Machine Learning (ICML'11), Bellevue, Washington, 2011. Note that the Matlab version of Libsvm [2] is needed (http://www.csie.ntu.edu.tw/~cjlin/libsvm/) to implement supervised SVM with different weights of labeled and unlabeled instances.It should be included to facilitate the implementation of S4VM. For your convenience, the matlab code of libsvm using different insatnce weights is included in the package. [2] C.-C. Chang and C.-J. Lin. Libsvm: a library for support vector machines, Department of Computer Science and Information Engineering, National Taiwan University, Taipei, Taiwan, Technical Report, 2001. You will find an example of using this code in the 'example.m' function. The example data is wdbc data. In particular, 10 examples are labeled and the rest are unlabeled. In our ICML'11 experiment, all the features are nomalized to [0,1] in advanced. ATTN: - This package is free for academic usage. You can run it at your own risk. For other purposes, please contact Prof. Zhi-Hua Zhou (zhouzh@nju.edu.cn). - This package was developed by Mr. Teng Zhang (zhangt@lamda.nju.edu.cn) and Mr.Yu-Feng Li (liyf@lamda.nju.edu.cn). For any problem concerning the code, please feel free to contact Mr. Li or Mr. Zhang. ------------------------------------------------------------------------------------------