A reliable and explainable framework for Random Forest
The software here contained is related to a scientific research providing a modelling and analysis framework for the boosting Random Forests for reliability and explainability. The framework is based on the formalism of Bayesian Networks.
The repository is structured in the following folders:
- Improving Classification: where the prototype of the paper cited below and related files are contained.
The software is licensed according to the GNU General Public License v3.0 (see License file).
- Stefano Marrone - UniversitĂ della Campania "Luigi Vanvitelli" (Italy)
This software is build upon the following software libraries; the most important of them are:
- Scikit-learn ver. 0.22.2.post1
- Pomegranate 0.13.5
Please refer to the following paper:
- de Biase, M.S., Marrone, S., Marulli, F., Verde, L.; Improving Classification Trustworthiness in Random Forests; submitted to 2021 IEEE CSR Workshop on Resilient Artificial Intelligence (RAI)