FN Archimer Export Format
PT J
TI Working group on machine learning in marine science (WGMLEARN; Outputs from 2021 meeting)
BT 
AF ICES
SI NANTES
   BREST
SE PDG-RBE-HALGO-EMH
   PDG-RBE-HALGO-LTBH
UM DECOD
TC 0
AC ICES
UR https://archimer.ifremer.fr/doc/00754/86646/92098.pdf
LA English
DT Article
AB The WGMLEARN group was formed to explore the use of machine learning in the marine sci-ences, and work towards increasing knowledge of and competence with relevant methods among marine scientists. The specific objectives were to review methods, applications, and im-plementations, to gather knowledge about them from a wide array of scientists, to address the implications of these methods for data management, and to highlight how they can be applied more/better in the future. To achieve those objectives, we performed an extensive literature survey, gathering around 900 published works, and categorized them to extract trends in the usage of methods or data types. Based on this, we drafted three manuscripts. The first describes the history of machine learning for marine ecology and highlights the domi-nance of images and acoustics as data sources, as well as the rise of deep learning methods. The second aims to guide new users towards these deep learning methods and, based on examples, shows their potential for a wide array of questions in marine sciences. The third focuses on ap-proaches that are of particular relevance for fisheries science and shows that machine learning can be relevant at all scales of fisheries studies. Overall, we recognize a continued need to accel-erate automation and effective data processing, and suggest new activities aimed at training, data management, infrastructure, and outreach, necessary to achieve this acceleration. 
PY 2022
SO ICES Scientific Reports/Rapports scientifiques du CIEM
SN 2618-1371
PU ICES
VL 4
IS 15
DI 10.17895/ices.pub.10060
ID 86646
ER

EF