Currently, the MALSAR package is available for MATLAB only. The software is licensed under the General Public License (GPL). The MALSAR package is free for academic use. If you use MALSAR in your research paper, please refer to the citation section for more information about citation. For any commercial use please contact us. By downloading the MALSAR package you agree to the terms and conditions of the license above.

View the Project on GitHub jiayuzhou/MALSAR

MALSAR: Multi-task learning via Structural Regularization

The MALSAR (Multi-tAsk Learning via StructurAl Regularization) package includes the following multi-task learning algorithms:

If you have any questions regarding MALSAR, please contact Jiayu Zhou at


The manual for the lastest version can be found here.


We gave a tutorial on multi-task learning at the Twelfth SIAM Internation conference on Data Mining (SDM'12). The tutorial slides can be found here.


In citing MALSAR in your papers, you can use the following:
Zhou, Jiayu, Jianhui Chen, and Jieping Ye. "MALSAR: Multi-task learning via structural regularization." Arizona State University (2011).

If you use LaTeX, you can use the BibTex entry:

  title={MALSAR: Multi-task learning via structural regularization},
  author={Zhou, Jiayu and Chen, Jianhui and Ye, Jieping},
  journal={Arizona State University},


The MALSAR software project has been supported by research grants from the National Science Foundation (NSF).