Since I started posting news on this website, I’ve come to greatly appreciate the projects coming from Dmitry Kazakov because they are always perfectly documented. The documentation written by Dmitry is hands down some of the best and most comprehensive I’ve ever encountered.
His new Fuzzy machine learning project is no different. In fact, in many ways he manages to top himself in the documentation department, which is no small feat.
I tip my hat to you Mr. Kazakov.
The project itself is:
Fuzzy machine learning framework is a library and a GUI front-end for machine learning using intuitionistic fuzzy data. The approach is based on the intuitionistic fuzzy sets and the possibility theory.
Actually it’s a whole lot more than that, but the space available here is limited. Instead I urge you to visit the project website for more information.
For those of you who’re not quite sure what this whole “fuzzy” thing is, Dmitry has made a paper available that talks in length about the subject:
The aim of this work is to propose an intuitionistic formulation of the machine learning problem, completely independent from probability theory. Intuitionistic fuzzy sets naturally appear in machine learning based on possibility theory as a result of uncertainty in the measuring process. Considering this only type of uncertainty, a notion of fuzzy feature as a mapping can be defined. Fuzzy features originate four different pattern spaces. Fuzzy classification is defined as a pair of images in the pattern spaces. Properties of the fuzzy pattern spaces required for fuzzy inference are demonstrated.
I think I just felt something fly over my head there. Fast.
You can read the full release announcement here.