Abstract
Support vector machine (SVM), which is analytically solved to reach its optimal structural formula, can be represented as a network architecture resembling artificial neural networks (multilayer perceptrons) that have been pruned to obtain model parsimony or improve generalization. This study presents two methods of pruning the support vector networks, and applies them to a case study of real-time flood stage forecasting. One method prunes the input variables by the cross-correlation method, while the other prunes the support vectors according to the derived relationship among SVM parameters. These pruning methods do not revise the SVM algorithm, and the pruned models therefore still have the optimal architecture. The real-time forecasting performance pertaining to the original and the pruned SVM models are compared, and comparison results indicate that the pruning reduces the network complexity but does not degrade the forecasting ability. Moreover, the deletion of support vectors and the change in their weights during the pruning process are identified, revealing that the support vectors with small weight (and hence less significant) tend to be pruned off, while the support vectors that are more informative to characterize the floods are preserved after the pruning. Finally, this study suggests that the proposed support vector pruning method be a potential data mining technique.
| Original language | English |
|---|---|
| Pages (from-to) | 67-78 |
| Number of pages | 12 |
| Journal | Journal of Hydrology |
| Volume | 347 |
| Issue number | 1-2 |
| DOIs | |
| Publication status | Published - 2007 Dec 15 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Water Science and Technology
Cite this
}
Pruning of support vector networks on flood forecasting. / Chen, Shien Tsung; Yu, Pao-Shan.
In: Journal of Hydrology, Vol. 347, No. 1-2, 15.12.2007, p. 67-78.Research output: Contribution to journal › Article
TY - JOUR
T1 - Pruning of support vector networks on flood forecasting
AU - Chen, Shien Tsung
AU - Yu, Pao-Shan
PY - 2007/12/15
Y1 - 2007/12/15
N2 - Support vector machine (SVM), which is analytically solved to reach its optimal structural formula, can be represented as a network architecture resembling artificial neural networks (multilayer perceptrons) that have been pruned to obtain model parsimony or improve generalization. This study presents two methods of pruning the support vector networks, and applies them to a case study of real-time flood stage forecasting. One method prunes the input variables by the cross-correlation method, while the other prunes the support vectors according to the derived relationship among SVM parameters. These pruning methods do not revise the SVM algorithm, and the pruned models therefore still have the optimal architecture. The real-time forecasting performance pertaining to the original and the pruned SVM models are compared, and comparison results indicate that the pruning reduces the network complexity but does not degrade the forecasting ability. Moreover, the deletion of support vectors and the change in their weights during the pruning process are identified, revealing that the support vectors with small weight (and hence less significant) tend to be pruned off, while the support vectors that are more informative to characterize the floods are preserved after the pruning. Finally, this study suggests that the proposed support vector pruning method be a potential data mining technique.
AB - Support vector machine (SVM), which is analytically solved to reach its optimal structural formula, can be represented as a network architecture resembling artificial neural networks (multilayer perceptrons) that have been pruned to obtain model parsimony or improve generalization. This study presents two methods of pruning the support vector networks, and applies them to a case study of real-time flood stage forecasting. One method prunes the input variables by the cross-correlation method, while the other prunes the support vectors according to the derived relationship among SVM parameters. These pruning methods do not revise the SVM algorithm, and the pruned models therefore still have the optimal architecture. The real-time forecasting performance pertaining to the original and the pruned SVM models are compared, and comparison results indicate that the pruning reduces the network complexity but does not degrade the forecasting ability. Moreover, the deletion of support vectors and the change in their weights during the pruning process are identified, revealing that the support vectors with small weight (and hence less significant) tend to be pruned off, while the support vectors that are more informative to characterize the floods are preserved after the pruning. Finally, this study suggests that the proposed support vector pruning method be a potential data mining technique.
UR - http://www.scopus.com/inward/record.url?scp=35748941976&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=35748941976&partnerID=8YFLogxK
U2 - 10.1016/j.jhydrol.2007.08.029
DO - 10.1016/j.jhydrol.2007.08.029
M3 - Article
VL - 347
SP - 67
EP - 78
JO - Journal of Hydrology
JF - Journal of Hydrology
SN - 0022-1694
IS - 1-2
ER -