Vector representations of the Canadian Geoscience Foundation model / M. Parsa. : M183-2/9318E-PDF

"The Geological Survey of Canada (GSC) has recently developed several data-driven, geospatial predictive models, with more expected to follow. Nevertheless, several challenges hinder the advancement of this initiative. First, the geoscientific phenomena, such as mineralization, targeted by geospatial predictive modelling tasks are typically rare events, which limits the effectiveness of conventional data-driven algorithms. Second, training models from scratch for each task is both resource-intensive and time-consuming. The Canadian Geoscience Foundation model (CGF) offers a promising solution to the aforementioned challenges by providing a general-purpose, adaptable AI model trained on extensive, diverse, pan-Canadian geoscientific datasets"--Page 1.

Lien permanent pour cette publication :
publications.gc.ca/pub?id=9.957831&sl=1

• Format MARC XML
• Format MARC HTML

Demander des formats alternatifs