is a framework especially designed for the benchmarking of feature selection algorithms and cost functions. Through a clean object-oriented structure and coded in C++, this framework allows the user to easily include new algorithms and/or cost functions and to carry out fast computational experiments. Moreover, featsel
allows the user to take advantage of the fact that the search space of a feature selection procedure can be organized into a Boolean lattice; to this end, several methods are included in this framework especially for Boolean manipulation of feature subsets. featsel
is also accompanied by several scripts to assist the user in the inclusion and removal of algorithms and cost functions, to carry out benchmarking experiments and to create tables and graphs with the obtained results.
framework architecture was firstly presented by Reis, Estrela et al. (2017)
, and was used to generate results of the computational experiments presented in Reis et al. (2019)
. Currently, it is being used in the project "Development of ABC-based approaches to tackle the lack of isolation problem in signaling pathway
Source code license: GNU GPL-3.0
Source code availability: github.com/msreis/featsel