Tools
Learn fuzzy logic, then build with it.
A short path for newcomers, the papers that started it all, and the toolkits researchers actually use.
Fuzzy logic is easier to start than it sounds. Begin with one example and one toolkit.
Learn fuzzy
1
Get the idea
Play with the demo on our home page. A fuzzy set says how much something belongs, from 0 to 1.
2
Build one system
Follow the classic tipping example in scikit-fuzzy. It takes about an hour.
3
Read the classics
Three short papers shaped the whole field. They are below.
4
Make it yours
Pick a toolkit below and model something from your own work.
Classic papers
- Fuzzy setsL. A. Zadeh · 1965 · Information and Control 8(3)
- An experiment in linguistic synthesis with a fuzzy logic controllerE. H. Mamdani, S. Assilian · 1975 · International Journal of Man-Machine Studies 7(1)
- Fuzzy identification of systems and its applications to modeling and controlT. Takagi, M. Sugeno · 1985 · IEEE Transactions on Systems, Man, and Cybernetics SMC-15(1)
Toolkits
Free and open source unless noted.
- Python
scikit-fuzzy
- Best for
- Learning fuzzy logic and building Mamdani systems in Python, with membership function plots.
- Use it when
- You know Python and want to prototype a controller or follow a textbook example.
- Python
Simpful
- Best for
- Writing fuzzy rules as readable text and running Mamdani or Takagi-Sugeno inference.
- Use it when
- You want rules a non-programmer can read, or you need Sugeno-style outputs.
- Python
pyFUME
- Best for
- Learning fuzzy models automatically from a dataset.
- Use it when
- You have tabular data and want an interpretable Takagi-Sugeno model built from it.
- C++, Java, Python
FuzzyLite
- Best for
- Fast fuzzy controllers in C++, Java or Python.
- Use it when
- You need to ship a controller inside an application or device.
- MATLAB
Fuzzy Logic Toolbox
- Best for
- Designing, tuning and simulating fuzzy systems with apps and Simulink blocks.
- Use it when
- Your course or lab already uses MATLAB. Requires a MATLAB license.