ebsl 0.1.3

The ensemble binomial subjective logic framework

Introduction

The ebsl is a Python module implementing the Ensemble Binomial Subjective Logic framework, a binary ensemble classifier (based on subjective logic) that provides dynamic weighting.

The C++ extension is documented here.

Results of the testing scenario in a vehicular networks are in a separate repository.

Proof of Concepts

Some basic proof of concepts related to Binomial opinions and the EBSL classifier exist in the PoC directory.

Installation

The EBSL framework is installable using pip. Simply run pip install ebsl. The instructions below are for building from source.

Dependencies

• Python (>= 3.12)
• Numpy (>= 2.0)
• Pandas

You will also need a C++20 compliant compiler to compile the C++ extension and cmake.

Building using pip

Clone the repository and move into its directory:

git clone --recurse-submodules https://github.com/a-h-ismail/ebsl-ids.git
cd ebsl-ids

Then install using pip:

pip install .

The extension will be automatically built and installed globally. Adapt the command as needed for virtual environments.