If your are using any of the material below please cite the corresponding publication:
ODTF (One-class Decision Tree Fuzzyfier) is an algorithm that wraps a linear DT and establishes class-membership scores based on weighted distances to decision thresholds. ODTF is designed to implement more robust supervised learners based on making class boundaries fuzzy. The first ODTF version here available works with binary classification problems and numerical features (real, integer, binary).
Software, scripts, tools:
The Python version can be downloaded from the ODTF GitHub repository.
- Mar 2018 - Experiments with the NSL-KDD and UNSW-NB15 datasets for reproducibility: ODTF GitHub repository. These experiments only include ODTF and linear DT performances.
- May 2018 - Complete experiments with the NSL-KDD and UNSW-NB15 datasets for reproducibility (also including comparison with soft/fuzzy decision trees, SDT): ODTF, LDT and SDT complete comparison.
Example of ODTF scoring for 100 samples of the Fisher's Iris dataset (only two features and two dimensions are taken for the example).