Machine Learning Standards in the Wild. DOME Webinar on ML Recommendations and Applications
Watch a DOME Webinar on Machine Learning Best Practices & Recommendations on 24th September 2024
In recent years, there has been a substantial increase in scientific publications in journals publishing computational research such as ours utilising Machine Learning (ML). This represents a significant challenge for disseminating and assessing scientific research as the black box and often non-open source nature of ML products potentially making intelligibility even more challenging than traditional computational research. To address this challenge there have been a number of efforts creating checklists, reporting standards and submission guidelines in ML research. DOME is one such effort, providing a set of community-wide guidelines, recommendations and checklists, spanning these four areas (Data, Optimization, Model and Evaluation in Machine Learning) aiming to help establish standards of supervised machine learning validation in biology. The work was initially published in 2021 with more details on the DOME-ML page. This activity is part of the ELIXIR Machine Learning Focus Group, and GigaScience Press has participated in the working group and has experimented using the standard in our peer-review and publishing of ML papers.
On Tuesday 24th September 2024 at 13:00 CEST our Editor in Chief Scott Edmunds will be participating in an ELIXIR organised webinar covering DOME Recommendations; definition and applications. The goal of this webinar is to provide an initial overview of the DOME recommendations, as well as the services that have been implemented for capturing them in the DOME registry. Moreover, it will include some perspectives on their practical use in the context of a scientific community (and specifically the ELIXIR Intrinsically Disordered Proteins Community), as well as in the context of our journal, where the DOME recommendations have already been adopted as part of peer review and publication process for papers using ML (see the slides of Chris Armit’s recent BOSC talk for a preview).
Chris Armit presenting on our DOME-ML trial at BOSC2024 in July 2024.
- Welcome
- What are the DOME recommendations? (Fotis Psomopoulos, INAB and Barbara Szomolay, Cardiff University)
- What is the DOME registry? (Silvio Tosatto, University of Padua (UNIPD))
- DOME use within ELIXIR – the case of the IDP Community (Damiano Piovesan, University of Padua (UNIPD))
- DOME use beyond ELIXIR – a journal’s perspective (Scott C. Edmunds, GigaScience Press)
- Final Q&A
- Fotis Psomopoulos, Institute of Applied Biosciences (INAB), Center for Research and Technology Hellas (CERTH)
- Barbara Szomolay, Cardiff University
- Silvio Tosatto, University of Padua (UNIPD)
- Damiano Piovesan, University of Padua (UNIPD)
Scott C. Edmunds, GigaScience Press
If you are interested in learning more you can register via this link and the recording of this event will be made available on the ELIXIR Europe Youtube channel.
References
Walsh I, Fishman D, Garcia-Gasulla D, Titma T, Pollastri G; ELIXIR Machine Learning Focus Group; Harrow J, Psomopoulos FE, Tosatto SCE. DOME: recommendations for supervised machine learning validation in biology. Nat Methods. 2021 Oct;18(10):1122-1127. doi:10.1038/s41592-021-01205-4.
