This package is a MATLAB implementation of ["Handling Multi-Class Problem by Intuitionistic Fuzzy Twin Support Vector Machines Based on Relative Density Information"]. You can train the M-IFTSVM-RD model on your dataset for multi-classification problems.
The M-IFTSVM-RD approximates the probability density distribution in multi-dimensional continuous space by computing the K-nearest-neighbor distance of each training sample.
In matlab, run M-IFTSVM-RD.m.
You need to optimize the parameter first, then run the code for the optimized parameter.
For the final result, you should consider MaxAcc.
ATTN1: These packages are free for academic usage. You can run them at your own risk. For other purposes.
This code belongs to the following paper. Please cite it if you want to use it:
S. Rezvani and J. Wu, "Handling Multi-Class Problem by Intuitionistic Fuzzy Twin Support Vector Machines Based on Relative Density Information," in IEEE Transactions on Pattern Analysis and Machine Intelligence, doi: 10.1109/TPAMI.2023.3310908.
[1] B. Gao, J. Wang, Y. Wang, and C. Yang, “Coordinate descent fuzzy twin support vector machine for classification,” in Proc. IEEE 14th Int. Conf. Mach. Learn. Appl., Dec. 2015, pp. 7–12.