In this work we propose new ensemble methods for the hierarchical classification of gene functions. Our methods exploit the hierarchical relationships between the classes in different ways: each ensemble node is trained “locally”, according to its position in the hierarchy; moreover, in the evaluation phase the set of predicted annotations is built so to minimize a global loss function defined over the hierarchy. We also address the problem of sparsity of annotations by introducing a cost- sensitive parameter that allows to control the precision-recall trade-off. Experiments with the model organism S. cerevisiae, using the FunCat taxonomy and 7 biomolecular data sets, reveal a significant advantage of our techniques over “flat” and cost-insensitive hierarchical ensembles.
Hierarchical cost-sensitive algorithms for genome-wide gene function prediction / N. Cesa Bianchi, G. Valentini - In: Machine learning in systems biology : proceedings of the third international workshop : september 5-6, 2009, Ljubljana, Slovenia / [a cura di] S. Dzeroski, P. Geurts, J. Rousu. - Helsinki : University of Helsinky, Department of Computer Science, 2009. - ISBN 9789521056994. - pp. 25-34 (( Intervento presentato al 3. convegno International Workshop on Machine Learning in Systems Biology tenutosi a Ljubljana, Slovenia nel 2009.
Titolo: | Hierarchical cost-sensitive algorithms for genome-wide gene function prediction | |
Autori: | CESA BIANCHI, NICOLO' ANTONIO (Primo) VALENTINI, GIORGIO (Ultimo) | |
Settore Scientifico Disciplinare: | Settore INF/01 - Informatica | |
Progetto: | Pattern Analysis, Statistical Modelling and Computational Learning 2 | |
Data di pubblicazione: | 2009 | |
URL: | http://mlsb09.ijs.si/files/MLSB09-Proceedings.pdf | |
Tipologia: | Book Part (author) | |
Appare nelle tipologie: | 03 - Contributo in volume |
File in questo prodotto:
File | Descrizione | Tipologia | Licenza | |
---|---|---|---|---|
workshop 2009.pdf | Publisher's version/PDF | Open Access Visualizza/Apri |