This analysis demonstrates improved predictions of cod stock dynamics using neural networks and environmental data, indicating enhanced fisheries management strategies.
Key Points
The aim is to predict Barents Sea cod stock dynamics using oceanographic data and neural networks.
Utilized lagged hydrographic variables as inputs for neural network models.
Combined diverse updated oceanographic and fishing data for analysis.
Applied machine learning techniques to model stock dynamics.
Neural networks provided more accurate predictions compared to traditional linear regression models.
Highlighted the importance of environmental variability on fish stock dynamics.
Results underscore the potential of machine learning in fisheries science.