期刊
COMPUTERS & GEOSCIENCES
卷 41, 期 -, 页码 126-135出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.cageo.2011.08.023
关键词
Probabilistic; Nonlinear; Artificial neural network; Interactions; Prediction interval; R programming language
A conditional density estimation network (CDEN) is a probabilistic extension of the standard multilayer perceptron neural network (MLP). A CDEN model allows users to estimate parameters of a specified probability density function conditioned upon values of a set of predictors using the MLP architecture. The result is a flexible model for the mean, the variance, exceedance probabilities, prediction intervals, etc. from the specified conditional distribution. Because the CDEN is based on the MLP, nonlinear relationships, including those involving complicated interactions between predictors, can be described by the modeling framework. CDEN models have been applied to a wide range of environmental prediction tasks, such as precipitation downscaling, extreme value analysis in hydrology, wind retrievals from satellites, and air quality forecasting. This paper describes the CaDENCE (Conditional Density Estimation Network Creation and Evaluation) package, which provides routines for creating and evaluating CDEN models in the R programming language. CaDENCE routines are demonstrated on a dataset consisting of suspended sediment concentrations and discharge measurements from the Fraser River at Hope, British Columbia, Canada. Crown Copyright (C) 2011 Published by Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据