Ensembles of genetically trained artificial neural networks for survival analysis
(2013) 21st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013 p.333-338- Abstract
We have developed a prognostic index model for survival data based on an ensemble of artificial neural networks that optimizes directly on the concordance index. Approximations of the c-index are avoided with the use of a genetic algorithm, which does not require gradient information. The model is compared with Cox proportional hazards (COX) and three support vector machine (SVM) models by Van Belle et al. [10] on two clinical data sets, and only with COX on one artificial data set. Results indicate comparable performance to COX and SVM models on clinical data and superior performance compared to COX on non-linear data.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/34253bab-8f63-432a-9ed5-f9f254e46752
- author
- Kalderstam, Jonas LU ; Edén, Patrik LU and Ohlsson, Mattias LU
- organization
- publishing date
- 2013
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- ESANN 2013 proceedings, 21st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
- pages
- 6 pages
- conference name
- 21st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013
- conference location
- Bruges, Belgium
- conference dates
- 2013-04-24 - 2013-04-26
- external identifiers
-
- scopus:84887057609
- ISBN
- 9782874190810
- language
- English
- LU publication?
- yes
- id
- 34253bab-8f63-432a-9ed5-f9f254e46752
- date added to LUP
- 2017-05-23 09:44:08
- date last changed
- 2024-01-13 21:32:56
@inproceedings{34253bab-8f63-432a-9ed5-f9f254e46752, abstract = {{<p>We have developed a prognostic index model for survival data based on an ensemble of artificial neural networks that optimizes directly on the concordance index. Approximations of the c-index are avoided with the use of a genetic algorithm, which does not require gradient information. The model is compared with Cox proportional hazards (COX) and three support vector machine (SVM) models by Van Belle et al. [10] on two clinical data sets, and only with COX on one artificial data set. Results indicate comparable performance to COX and SVM models on clinical data and superior performance compared to COX on non-linear data.</p>}}, author = {{Kalderstam, Jonas and Edén, Patrik and Ohlsson, Mattias}}, booktitle = {{ESANN 2013 proceedings, 21st European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning}}, isbn = {{9782874190810}}, language = {{eng}}, pages = {{333--338}}, title = {{Ensembles of genetically trained artificial neural networks for survival analysis}}, year = {{2013}}, }