Conditional Forecasting of Water Level Time Series with RNNs

Bart van der Lugt, A.J. Feelders

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

We describe a practical situation in which the application of forecasting models could lead to energy efficiency and decreased risk in water level management. The practical challenge of forecasting water levels in the next 24 h and the available data are provided by a dutch regional water authority. We formalized the problem as conditional forecasting of hydrological time series: the resulting models can be used for real-life scenario evaluation and decision support. We propose the novel Encoder/Decoder with Exogenous Variables RNN (ED-RNN) architecture for conditional forecasting with RNNs, and contrast its performance with various other time series forecasting models. We show that the performance of the ED-RNN architecture is comparable to the best performing alternative model (a feedforward ANN for direct forecasting), and more accurately captures short-term fluctuations in the water heights.
Original languageEnglish
Title of host publicationAdvanced Analytics and Learning on Temporal Data
Subtitle of host publication4th ECML PKDD Workshop, AALTD 2019, Würzburg, Germany, September 20, 2019, Revised Selected Papers
EditorsVincent Lemaire, Simon Malinowski, Anthony Bagnall, Alexis Bondu, Thomas Guyet, Romain Tavenard
PublisherSpringer
Pages55-71
ISBN (Electronic)978-3-030-39098-3
ISBN (Print)978-3-030-39097-6
DOIs
Publication statusPublished - 2020
EventWorkshop on Advanced Analytics and Learning on Temporal Data - Wurzburg, Germany
Duration: 20 Sept 201920 Sept 2019
https://project.inria.fr/aaltd19/

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume11986

Workshop

WorkshopWorkshop on Advanced Analytics and Learning on Temporal Data
Abbreviated titleAALTD19
Country/TerritoryGermany
CityWurzburg
Period20/09/1920/09/19
Internet address

Keywords

  • Time series
  • Conditional forecasting
  • Encoder/Decoder
  • Exogenous variables
  • Recurrent Neural Network

Fingerprint

Dive into the research topics of 'Conditional Forecasting of Water Level Time Series with RNNs'. Together they form a unique fingerprint.

Cite this