Welcome to MDFS home!


MDFS is a software designed and implemented to assist in MultiDimensional Feature Selection (MDFS).


MDFS is released under GPLv3.


The engine core is written in optimised C++ and CUDA code.

We currently offer an R package, available from CRAN, as well as a Python package, available from PyPI. Their versions stay in sync.


The changelog is available here.

Learn more

Want to know more? We have published two papers discussing both the approach and the software itself:

  1. K. Mnich and W. R. Rudnicki, ‘All-relevant feature selection using multidimensional filters with exhaustive search’, Information Sciences, vol. 524, pp. 277–297, Jul. 2020, doi: 10.1016/j.ins.2020.03.024.
  2. R. Piliszek et al., ‘MDFS: MultiDimensional Feature Selection in R’, The R Journal, vol. 11, no. 1, pp. 198–210, 2019, doi: 10.32614/RJ-2019-019.

© 2023 Radosław Piliszek, Krzysztof Mnich, Witold Rudnicki and contributors