/Graded-Relations-From-Data

Implementation of the framework in the paper: Waegeman, W., Pahikkala, T., Airola, A., Salakoski, T., Stock, M., & De Baets, B. (2012). A kernel-based framework for learning graded relations from data. IEEE Transactions on Fuzzy Systems, 20(6), 1090-1101.

Primary LanguagePythonMIT LicenseMIT

Graded Relations from Data

Implementation of the framework in the paper:

Waegeman, W., Pahikkala, T., Airola, A., Salakoski, T., Stock, M., & De Baets, B. (2012). A kernel-based framework for learning graded relations from data. IEEE Transactions on Fuzzy Systems, 20(6), 1090-1101.

Setup

Prerequisites

  • Python 3.6+
  • pip (for setting up dependencies)

Setting up dependencies

Use

$ pip install -r requirements.txt

Setting up the package

Use

$ python setup.py install

Running the experiments

The following describe the procedure to run the experiments in VI. A. and VI. C. (including generating simulated data) of the paper.

VI. A.

Use

$ python scripts/experiment1.py [options]

To view the list of available options and their descriptions, use

$ python scripts/experiment1.py --help

VI. C.

Use

$ python scripts/experiment2.py [options]

To view the list of available options and their descriptions, use

$ python scripts/experiment2.py --help

License

This code is provided using the MIT License.

This project was a part of the course MA6040: Fuzzy Logic Connectives: Theory and Applications, offered in Spring 2019 at IIT Hyderabad.

Team members: Vishwak Srinivasan and Sukrut Rao.