Structural Similarity Support Vector Machines

This software package adds structural similarity (through neighborhood vectors) into linear or histogram intersection kernel SVM. This method is particularly effective in dealing with imbalanced learning problems.

For technical details, please refer to the following paper in the publication page:

Balance Support Vector Machines Locally Using the Structural Similarity Kernel [pdf]
Jianxin Wu
PAKDD 2011 : The 15th Pacific-Asia Conference on Knowledge Discovery and Data Mining

The code is available here.

Please read the README file before you use this software.

If you use this software, please cite the above paper.

Contact: Jianxin Wu (wujx2001 AT gmail.com)