search

ices-eg / DIG

ICES Data and Information Group
Creative Commons Zero v1.0 Universal
3 stars 1 forks source link

Lack of familiarisation with Machine Learning techniques and tools causes less efficiency and risks to scientific integrity #511

Open sjurl opened 1 year ago

sjurl commented 1 year ago
Aspect Description
Challenges Few experts in the ICES community, but there is a working group now. On the short term there is too few experts, but in the long run there will be more. WGMLEARN has been out of commission for a couple of years.
Action required ​​DIG to Monitor. Follow up on the WGMLEARN.
Ownership No specific owner

Linked to Opportunity #581

legacy entry

OliverWilliamsDataManager commented 10 months ago

This has been converted from the original issue to cover challenges only, the opportunities are better covered by other working groups (especially the ML one mentioned above).

I deemed the impact medium because although ML can exploit data in new ways, the data collection, curation, standardisation and sharing has to come before this is possible.

I deemed the likelihood as medium because ML outputs will probably become one of the important data streams for ICES but traditional streams will continue to be crucial as well.