Τίτλος:
Hydrological post-processing using stacked generalization of quantile regression algorithms: Large-scale application over CONUS
Γλώσσες Τεκμηρίου:
Αγγλικά
Περίληψη:
Post-processing of hydrological model simulations using machine learning algorithms can be applied to quantify the uncertainty of hydrological predictions. Combining multiple diverse machine learning algorithms (referred to as base-learners) using stacked generalization (stacking, i.e. a type of ensemble learning) is considered to improve predictions relative to the base-learners. Here we propose stacking of quantile regression and quantile regression forests. Stacking is performed by minimising the interval score of the quantile predictions provided by the ensemble learner, which is a linear combination of quantile regression and quantile regression forests. The proposed ensemble learner post-processes simulations of the GR4J hydrological model for 511 basins in the contiguous US. We illustrate its significantly improved performance relative to the base-learners used and a less prominent improvement relative to the “hard to beat in practice” equal-weight combiner. © 2019 Elsevier B.V.
Συγγραφείς:
Tyralis, H.
Papacharalampous, G.
Burnetas, A.
Langousis, A.
Περιοδικό:
Journal of Hydrologic Engineering
Λέξεις-κλειδιά:
Climate models; Forecasting; Forestry; Hydrology; Machine learning; Regression analysis, Ensemble learning; Hydrological uncertainty; Interval score; Probabilistic forecasts; Quantile regression, Learning algorithms, algorithm; forecasting method; hydrological modeling; machine learning; simulation; stacking; uncertainty analysis, United States
DOI:
10.1016/j.jhydrol.2019.123957
