Page 32 - Crowdsourcing AI and Machine Learning solutions for SDGs - ITU AI/ML Challenges 2024 Report

Crowdsourcing AI and Machine Learning solutions for SDGs

11 Challenge Solution Contributions

11.1 Standards

ITU has developed a range of standards-based Machine Learning mechanisms in 5G. The goal
is to provide a full toolkit to build Machine Learning into networks. Participants of the ITU AI/
ML Challenge, especially communication networks, are encouraged to base their work on ITU
standards.

Through the challenge engagements, participants were able to make contributions and
submissions that were used by the FG-ML5G and FG-AN to improve deliverables that were
under development as part of pre-standardization activities. These focus group deliverables
have now turned into ITU recommendations. Examples include the Y.3172 series as an example
of the link between standards and community and mentoring as well as Y.3061 as an example
of the link between open source, build-a-thon, and community and mentoring.

11.2 Open Source

The Challenge encourages the submission of open-source implementations, based on ITU
standards. Open-source implementations will enable a broad range of stakeholders to access
the outcomes of the Challenge and continue collaborating with relevant Challenge participants.
Participants are encouraged to submit on the challenge GitHub code, report, slides, demos,
publications, and any other supporting materials.

The following are links to various GitHub repos of the ITU AI/ML Challenge

• AI/ML in 5G Challenge & tinyML: https:// github .com/ ITU -AI -ML -in -5G -Challenge
• GeoAI: https:// github .com/ ITU -GeoAI -Challenge
• Fusion Energy: https:// github .com/ AI -for -Fusion -Energy

However, solutions based on proprietary implementations are also accepted.

11.3 Journal and Conference Publications

The ITU AI/ML Challenge has facilitated the publication of numerous solutions in journals
and conferences, contributing to the body of knowledge in the field of artificial intelligence
and machine learning. Participants' work has been recognized and disseminated through
platforms such as the ITU Journal on Future and Evolving Technologies (ITU J-FET), IEEE
journals, and ACM conferences. Below is an overview of the publication activities and special
issues dedicated to AI and ML for communication networks.

As of December 2023, thirty-six (36) papers have been published by ITU J-FET and the fourth
special edition of the journal is going through the review process. The call for papers of the
ITU J-FET is included in Figure 16 below.

24