DocumentCode :
2532061
Title :
Learning ensembles of Continuous Bayesian Networks: An application to rainfall prediction
Author :
Hellman, Scott ; McGovern, Amy ; Xue, Ming
Author_Institution :
Sch. of Comput. Sci., Univ. of Oklahoma, Norman, OK, USA
fYear :
2012
fDate :
24-26 Oct. 2012
Firstpage :
112
Lastpage :
117
Abstract :
We introduce Ensembled Continuous Bayesian Networks (ECBN), an ensemble approach to learning salient dependence relationships and to predicting values for continuous data. By training individual Bayesian networks on both a subset of the data (bagging) and a subset of the attributes in the data (randomization), ECBN produces models for continuous domains that can be used to identify important variables in a dataset and to identify relationships between those variables. We use linear Gaussian distributions within our ensembles, providing efficient network-level inference. By ensembling these networks, we are able to represent nonlinear relationships. We empirically demonstrate that ECBN outperforms the meteorological forecast on a rainfall prediction task across the United States, and performs comparably to results reported for Random Forests.
Keywords :
Gaussian distribution; belief networks; geophysics computing; inference mechanisms; learning (artificial intelligence); rain; ECBN; ensembled continuous Bayesian networks; learning ensembles; linear Gaussian distributions; meteorological forecast; network-level inference; nonlinear relationships; rainfall prediction; random forests; salient dependence relationships; Bagging; Bayesian methods; Educational institutions; Predictive models; Random variables; Training data; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Data Understanding (CIDU), 2012 Conference on
Conference_Location :
Boulder, CO
Print_ISBN :
978-1-4673-4625-2
Type :
conf
DOI :
10.1109/CIDU.2012.6382191
Filename :
6382191
Link To Document :
بازگشت