Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

Nagendra Kayastha; Dmitri Solomatine; D. Lal Shrestha

Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

Nagendra Kayastha, Dmitri Solomatine, D. Lal Shrestha

Water Resources

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

29 Downloads (Pure)

Abstract

In the MLUE method (reported in Shrestha et al. [1, 2]) we run a hydrological model M for multiple realizations of parameters vectors (Monte Carlo simulations), and use this data to build a machine learning model V to predict uncertainty (quantiles) of the model M output. In this paper, for model V, we employ three machine learning techniques, namely, artificial neural networks, model tree, locally weighted regression which leads to several models results. We propose to use the simple averaging method (SA) and the weighted model averaging method (WMA) to form a committee of these models. These approaches are applied to estimate uncertainty of streamflows simulation in Bagmati catchment in Nepal. Tests on the different data sets show that WMA performs a bit better than SA.

Original language	English
Title of host publication	Proceedings of the HIC 2014 - 11th international conference on hydroinformatics
Editors	M Piasecki
Pages	2364-2368
Publication status	Published - 2014
Event	11th International Conference on Hydroinformatics - New York, United States Duration: 17 Aug 2014 → 21 Aug 2014 Conference number: 11 http://academicworks.cuny.edu/cc_conf_hic/

Publication series

Name
Publisher	s.n.

Conference

Conference	11th International Conference on Hydroinformatics
Abbreviated title	HIC 2014
Country/Territory	United States
City	New York
Period	17/08/14 → 21/08/14
Internet address	http://academicworks.cuny.edu/cc_conf_hic/

Keywords

uncertainty analysis
hydrological model
machine learning
MLUE
model averaging

Access to Document

viewcontent.cgi8Final published version, 317 KBLicence: CC BY

Cite this

@inproceedings{f019e42bf46e4371a4923fed87e04ce2,

title = "Prediction Of Hydrological Models{\textquoteright} Uncertainty By A Committee Of Machine Learning-Models",

abstract = "In the MLUE method (reported in Shrestha et al. [1, 2]) we run a hydrological model M for multiple realizations of parameters vectors (Monte Carlo simulations), and use this data to build a machine learning model V to predict uncertainty (quantiles) of the model M output. In this paper, for model V, we employ three machine learning techniques, namely, artificial neural networks, model tree, locally weighted regression which leads to several models results. We propose to use the simple averaging method (SA) and the weighted model averaging method (WMA) to form a committee of these models. These approaches are applied to estimate uncertainty of streamflows simulation in Bagmati catchment in Nepal. Tests on the different data sets show that WMA performs a bit better than SA.",

keywords = "uncertainty analysis, hydrological model, machine learning, MLUE, model averaging",

author = "Nagendra Kayastha and Dmitri Solomatine and {Lal Shrestha}, D.",

year = "2014",

language = "English",

isbn = "978-1-5108-0039-7",

publisher = "s.n.",

pages = "2364--2368",

editor = "M Piasecki",

booktitle = "Proceedings of the HIC 2014 - 11th international conference on hydroinformatics",

note = "11th International Conference on Hydroinformatics, HIC 2014 ; Conference date: 17-08-2014 Through 21-08-2014",

url = "http://academicworks.cuny.edu/cc_conf_hic/",

}

Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models. / Kayastha, Nagendra; Solomatine, Dmitri; Lal Shrestha, D.
Proceedings of the HIC 2014 - 11th international conference on hydroinformatics. ed. / M Piasecki. 2014. p. 2364-2368.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

AU - Kayastha, Nagendra

AU - Solomatine, Dmitri

AU - Lal Shrestha, D.

N1 - Conference code: 11

PY - 2014

Y1 - 2014

N2 - In the MLUE method (reported in Shrestha et al. [1, 2]) we run a hydrological model M for multiple realizations of parameters vectors (Monte Carlo simulations), and use this data to build a machine learning model V to predict uncertainty (quantiles) of the model M output. In this paper, for model V, we employ three machine learning techniques, namely, artificial neural networks, model tree, locally weighted regression which leads to several models results. We propose to use the simple averaging method (SA) and the weighted model averaging method (WMA) to form a committee of these models. These approaches are applied to estimate uncertainty of streamflows simulation in Bagmati catchment in Nepal. Tests on the different data sets show that WMA performs a bit better than SA.

AB - In the MLUE method (reported in Shrestha et al. [1, 2]) we run a hydrological model M for multiple realizations of parameters vectors (Monte Carlo simulations), and use this data to build a machine learning model V to predict uncertainty (quantiles) of the model M output. In this paper, for model V, we employ three machine learning techniques, namely, artificial neural networks, model tree, locally weighted regression which leads to several models results. We propose to use the simple averaging method (SA) and the weighted model averaging method (WMA) to form a committee of these models. These approaches are applied to estimate uncertainty of streamflows simulation in Bagmati catchment in Nepal. Tests on the different data sets show that WMA performs a bit better than SA.

KW - uncertainty analysis

KW - hydrological model

KW - machine learning

KW - MLUE

KW - model averaging

UR - http://resolver.tudelft.nl/uuid:f019e42b-f46e-4371-a492-3fed87e04ce2

M3 - Conference contribution

SN - 978-1-5108-0039-7

SP - 2364

EP - 2368

BT - Proceedings of the HIC 2014 - 11th international conference on hydroinformatics

A2 - Piasecki, M

T2 - 11th International Conference on Hydroinformatics

Y2 - 17 August 2014 through 21 August 2014

ER -

Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this