Probabilistic Forecasting Using Monte Carlo Dropout Neural Networks

Cristián Serpell, IGNACIO DANIEL ARAYA ZAMORANO, Carlos Valle, Héctor Allende

Resultado de la investigación: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

3 Citas (Scopus)

Resumen

Using artificial neural networks for forecasting tasks is a popular approach that has proven to be very accurate. When used to estimate prediction intervals, a normal distribution is usually assumed as the data noise uncertainty term, as in MVE networks, while model parameters uncertainty is often ignored. Because of this, prediction intervals estimated by them are narrow in uncertain regions where train data is scarce. To tackle this problem we apply Monte Carlo dropout, which is a model uncertainty representation technique, to the network parameters of a Long Short-Term Memory MVE network, allowing us to construct better prediction intervals in probabilistic forecasting tasks. We compare our proposal with the pure MVE method in four wind speed and one consumer load real forecasting datasets, showing that our method improves results in terms of the Winkler loss in both one step ahead and multi-step ahead probabilistic forecasting.

Idioma originalInglés
Título de la publicación alojadaProgress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 24th Iberoamerican Congress, CIARP 2019, Proceedings
EditoresIngela Nyström, Yanio Hernández Heredia, Vladimir Milián Núñez
EditorialSpringer
Páginas387-397
Número de páginas11
ISBN (versión impresa)9783030339036
DOI
EstadoPublicada - 2019
Publicado de forma externa
Evento24th Iberoamerican Congress on Pattern Recognition, CIARP 2019 - Havana, Cuba
Duración: 28 oct 201931 oct 2019

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen11896 LNCS
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conferencia

Conferencia24th Iberoamerican Congress on Pattern Recognition, CIARP 2019
País/TerritorioCuba
CiudadHavana
Período28/10/1931/10/19

Huella

Profundice en los temas de investigación de 'Probabilistic Forecasting Using Monte Carlo Dropout Neural Networks'. En conjunto forman una huella única.

Citar esto