To Be Fair: A fair Benchmark of Group-Contribution and Machine-Learning based Property Models
Explore the screenshots »

Report a Bug · Request a Feature . Ask a Question

Table of Contents

• About
  • Built With
• Getting Started
  • Prerequisites
  • Installation
• Usage
• Roadmap
• Support
• Project assistance
• Contributing
• Authors & contributors
• Security
• License
• Acknowledgements

---

About

In every ML-based property prediction paper, classical Group-contribution (GC) models have been a constant target for scrutiny due to the workflow associated with their development. A main point that is constantly highlighted is the fact that their ability to extrapolate remains untested considering that all available data were used for model calibration, leaving none for validation. While some have advocated that this approach is not necessarily wrong since classical GC models are parametric and thus the risk of overfitting is small, their ability to extrapolate to data beyond the training set is still not verified. In this work, we develop a framework for fair development and comparison of GC-based ML models for a wide range of properties. Screenshots Classical Approach Fair Approach

Built With

[?] Please provide the technologies that are used in the project.

Getting Started

Prerequisites

[?] What are the project requirements/dependencies?

Installation

[?] Describe how to install and get started with the project.

Usage

[?] How does one go about using it? Provide various use cases and code examples here.

Roadmap

See the open issues for a list of proposed features (and known issues).

• Top Feature Requests (Add your votes using the 👍 reaction)
• Top Bugs (Add your votes using the 👍 reaction)
• Newest Bugs

Support

[?] Provide additional ways to contact the project maintainer/maintainers.

Reach out to the maintainer at one of the following places:

• GitHub issues
• Contact options listed on this GitHub profile

Project assistance

If you want to say thank you or/and support active development of my_project:

• Add a GitHub Star to the project.
• Tweet about the my_project.
• Write interesting articles about the project on Dev.to, Medium or your personal blog.

Together, we can make my_project better!

Contributing

First off, thanks for taking the time to contribute! Contributions are what make the open-source community such an amazing place to learn, inspire, and create. Any contributions you make will benefit everybody else and are greatly appreciated.

Please read our contribution guidelines, and thank you for being involved!

Authors & contributors

The original setup of this repository is by Adem R.N. Aouichaoui.

For a full list of all authors and contributors, see the contributors page.

Security

my_project follows good practices of security, but 100% security cannot be assured. my_project is provided "as is" without any warranty. Use at your own risk.

For more information and to report security issues, please refer to our security documentation.

License

This project is licensed under the GPLv3.

See LICENSE for more information.

Acknowledgements

Some of the code used in this project is adapted from the work by Paul Seghers. The project was conducted as part of his M.Sc. thesis in 2024.

The original code can be found in the GraPE repository.