Gene profile classification is achieved by casting the classification problem as finding the sparse representation of testing samples with respect to training samples. The sparse representation is found by subspace pursuit, which is much more efficient than linear programming techniques. The new approach, with no need of model selection, however, still has the performance which can match the best result achieved among all the SVM variants after careful model selection.
CITATION STYLE
Hang, X., Dai, W., & Wu, F. X. (2009). Subspace pursuit for gene profile classificaiton. In 2009 IEEE International Workshop on Genomic Signal Processing and Statistics, GENSIPS 2009. https://doi.org/10.1109/GENSIPS.2009.5174349
Mendeley helps you to discover research relevant for your work.