/MJP

Implementation of "Maximum Joint Probability with Multiple Representations for Clustering", IEEE Transactions on Neural Networks and Learning Systems.

Primary LanguageMATLAB

Maximum Joint Probability with Multiple Representations for Clustering

This repository is our implementation of

Rui Zhang, Hongyuan Zhang, and X. Li, "Maximum Joint Probability With Multiple Representations for Clustering," IEEE Transactions on Neural Networks and Learning Systems, 2021.

The core idea is to learn weights of different views based on which view fits the prior distribution better. The framework starts from a Bayesian aspect and we show the connection between the proposed maximum joint probability and some existing clustering methods. In the following illustration, although the 2-th view is well-structured for clustering, it is not a preferable view as it does not fit the prior distribution, i.e., Gaussian distribution.

Figure

If you have any questions, pleas email:

hyzhang98@gmail.com or hyzhang98@mail.nwpu.edu.cn.

How to Run MJP

Please see the comments in the source code file.

Citation

@ARTICLE{MJP,
  author={R. {Zhang} and H. {Zhang} and X. {Li}},
  journal={IEEE Transactions on Neural Networks and Learning Systems}, 
  title={Maximum Joint Probability With Multiple Representations for Clustering}, 
  year={2021},
  volume={},
  number={},
  pages={1-11},
  doi={10.1109/TNNLS.2021.3056420}
}