Hyperbox-brain - A Python toolbox for hyperbox-based machine learning algorithms - open-source library description

Aug 3, 2026

Upload by Asher

Hyperbox-based machine learning algorithms support classifier construction through fuzzy set and logic theory, and they offer modern predictor strengths such as scalability, explainability, online adaptation, learning from limited data, missing-data handling, and the ability to incorporate new classes. However, no comprehensive benchmark package existed for easy access by non-experts. Hyperbox-brain is an open-source Python library providing a unified API compatible with scikit-learn and numpy, installable via PyPI/conda and distributed under GPL-3.

LanguageEnglish

CategoryTechnology

FormatPDF

Pages7

Reading Time18 min

AccessFree to download

Motivation and significance

Hyperbox representation and learning mechanism

Key capabilities for lifelong learning

Algorithm taxonomy and library coverage

| Original software publication
hyperbox-brain: A Python toolbox for hyperbox-based machine learning algorithms
Thanh Tung Khuat ∗, Bogdan Gabrys
Complex Adaptive Systems Lab, Data Science Institute, University of Technology Sydney, NSW 2007, Australia | | | | --- | --- | --- | | a r t i c l e i n f o | a b s t r a c t
Hyperbox-based machine learning algorithms are an important and popular branch of machine learning in the construction of classifiers using fuzzy sets and logic theory and neural network architectures. This type of learning is characterised by many strong points of modern predictors such as a high scalability, explainability, online adaptation, effective learning from a small amount of data, native ability to deal with missing data and accommodating new classes. Nevertheless, thereis no comprehensive existing package for hyperbox-based machine learning which can serve as a benchmark for research and allow non-expert users to apply these algorithms easily. The hyperboxbrain is an open-source Python library implementing the leading hyperbox-based machine learning algorithms. This library exposes a unified API which closely follows and is compatible with the renowned scikit-learn and numpy toolboxes. The library may be installed from Python Package Index (PyPI) and the conda package manager and is distributed under the GPL-3 license. The source code, documentation, detailed tutorials, and the full descriptions of the API are available at [https://uts](https://uts)[caslab.github.io/hyperbox-brain](caslab.github.io/hyperbox-brain).
© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ([http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/licenses/by/4.0/)). | | | Article history:
Received 4 November 2022
Received in revised form 28 May 2023 Accepted 29 May 2023 | | | | Keywords:
Hyperbox-based machine learning Hyperbox fuzzy sets
Fuzzy min–max neural networks General fuzzy min–max neural network Explainable machine learning Classifier | | | Code metadata | Current code version | v0.1.5 | | --- | --- | | Permanent link to code/repository used for this code version
Permanent link to Reproducible Capsule | [https://github.com/ElsevierSoftwareX/SOFTX-D-22-00360](https://github.com/ElsevierSoftwareX/SOFTX-D-22-00360) | | Legal Code License | GPL-3 | | Code versioning system used | git | | Software code languages, tools, and services used | Python | | Compilation requirements, operating environments & dependencies | Python ≥ 3.6, scikit-learn ≥ 0.24, numpy ≥ 1.14.6, scipy ≥ 1. 1, joblib ≥ 0. 11, threadpoolctl ≥ 2.0.0 pandas ≥ 0.25, matplotlib ≥ 2.2.3, plotly ≥ 4.10 | | If available Link to developer documentation/manual | [https://hyperbox-brain.readthedocs.io/](https://hyperbox-brain.readthedocs.io/) | | Support email for questions | [thanhtung09t2@gmail.com](thanhtung09t2@gmail.com) | 1. Motivation and significance The hyperbox-brain toolbox has been developed by the researchers within the Complex Adaptive Systems laboratory at the University Technology Sydney. It is a result of many years of developing versatile machine learning algorithms with hyperboxes as the foundational representation element at their core. ∗ Corresponding author. E-mail address: [thanhtung.khuat@uts.edu.au](thanhtung.khuat@uts.edu.au) (Thanh Tung Khuat). Hyperbox-based machine learning algorithms use min–max hyperboxes as their fundamental building blocks to partition the sample space into various regions. A collection of hyperboxes representing the same class can form the regions of arbitrary shape and complexity. Each min–max hyperbox is usually characterised by the minimum and maximum vertices together with a fuzzy membership function acting as a distance or similarity measure. During the training procedure, these hyperboxes are formed, as needed, and adjusted to accommodate the incoming input samples based on the degree-of-fi

FAQs

What problem does the hyperbox-brain toolbox address?

It provides a comprehensive open-source package and benchmark for hyperbox-based machine learning algorithms, enabling easier use by non-experts.

How does hyperbox-based learning represent data in hyperbox-brain?

It uses min–max hyperboxes as fundamental building blocks to partition the sample space, with fuzzy membership values measuring distance or similarity for training and adjustment.

Which learning properties does hyperbox-based machine learning emphasize?

Scalability, explainability, incremental adaptation, effective learning from limited data, handling new samples, absorbing new knowledge while reducing catastrophic forgetting, and managing the stability–plasticity dilemma.

Aug 3, 2026

Upload by Asher