This repository contains an implementation of the method for feature selection, demonstrated in [1].
Intuitionistic fuzzy sets are built for each attribute and, then, a quantitative characteristic is computed for further ranking of features. Detailed discription of the method and interpretation of the results is presented in ipynb file.
All the visualizations (including the plotly.express ones) are availiable in Google Colaboratory.
-
Szmidt, E., Kacprzyk, J., & Bujnowski, P. ł. (2019). A New Method for Attributes Selection in Intuitionistic Fuzzy Models. Proceedings of the 2019 Conference of the International Fuzzy Systems Association and the European Society for Fuzzy Logic and Technology (EUSFLAT 2019), 408–415. https://doi.org/10.2991/eusflat-19.2019.57
-
Szmidt, E., & Kacprzyk, J. (2010). Dealing with typical values via Atanassov’s intuitionistic fuzzy sets. International Journal of General Systems, 39(5), 489–506. https://doi.org/10.1080/03081079.2010.484268
-
Szmidt, E., & Baldwin, J. F. (2006). Intuitionistic Fuzzy Set Functions, Mass Assignment Theory, Possibility Theory and Histograms. 2006 IEEE International Conference on Fuzzy Systems, 35–41. https://doi.org/10.1109/fuzzy.2006.1681691