A mixture-density-network based approach for finding rating curves: Facing multi-modality and unbalanced data distribution

Chulsang Yoo, Jooyoung Park

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)

Abstract

In this paper, the use of MDNs (Mixture Density Networks) is proposed for deciding rating curves. This method is beneficial especially when a single curve is developed when the relation between stage and discharge exhibits hysteresis. The computational analyses performed for the Han River and Mokkye stations showed that the MDN-based method yields more meaningful results than the conventional least squares approach. Of particular significance was the possible identification of the bi-modal characteristics of rating curves under the proposed method.

Original languageEnglish
Pages (from-to)243-250
Number of pages8
JournalKSCE Journal of Civil Engineering
Volume14
Issue number2
DOIs
Publication statusPublished - 2010 Mar

Bibliographical note

Funding Information:
The implementation of the MDNs of this paper is based on the NETLAB software (Nabney, 2001) at http://neural-server.aston. ac.uk/. This work was supported partly by a grant (KRF-2008-313-D01083) from the Korea Research Foundation, also partly by a grant (Project No. KIWE2008-0003) from K-water. All contributions are gratefully acknowledged.

Keywords

  • Hysteresis
  • Mixture density networks
  • Multi-layer perceptrons
  • Multi-modality
  • Neural networks
  • Rating curves
  • Scaled conjugate gradients algorithms

ASJC Scopus subject areas

  • Civil and Structural Engineering

Fingerprint

Dive into the research topics of 'A mixture-density-network based approach for finding rating curves: Facing multi-modality and unbalanced data distribution'. Together they form a unique fingerprint.

Cite this