Optimization of deep learning precipitation models using categorical binary metrics

This work introduces a methodology for optimizing neural network models using a combination of continuous and categorical binary indices in the context of precipitation forecasting. Probability of detection or false alarm rate are popular metrics used in the verification of precipitation models. How...

Descripción completa

Detalles Bibliográficos
Autores: Larraondo, P.R., Renzullo, L.J., Van Dijk, A.I.J.M., Inza, I., Lozano, J.A.
Tipo de recurso: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2020
País:España
Institución:Basque Center for Applied Mathematics (BCAM)
Repositorio:BIRD. BCAM's Institutional Repository Data
OAI Identifier:oai:bird.bcamath.org:20.500.11824/1106
Acceso en línea:http://hdl.handle.net/20.500.11824/1106
https://doi.org/10.1002/qj.828.
Access Level:acceso abierto
Palabra clave:machine learning
optimization
deep learning
precipitations
binary metrics
Descripción
Sumario:This work introduces a methodology for optimizing neural network models using a combination of continuous and categorical binary indices in the context of precipitation forecasting. Probability of detection or false alarm rate are popular metrics used in the verification of precipitation models. However, machine learning models trained using gradient descent cannot be optimized based on these metrics, as they are not differentiable. We propose an alternative formulation for these categorical indices that are differentiable and we demonstrate how they can be used to optimize the skill of precipitation neural network models defined as a multi-objective optimization problem. To our knowledge, this is the first proposal of a methodology for optimizing weather neural network models based on categorical indices.