Principled Non-Linear Feature Selection
Abstract
Recent non-linear feature selection approaches employing greedy optimisation of Centred Kernel Target Alignment(KTA) exhibit strong results in terms of generalisation accuracy and sparsity. However, they are computationally prohibitive for large datasets. We propose randSel, a randomised feature selection algorithm, with attractive scaling properties. Our theoretical analysis of randSel provides strong probabilistic guarantees for correct identification of relevant features. RandSel's characteristics make it an ideal candidate for identifying informative learned representations. We've conducted experimentation to establish the performance of this approach, and present encouraging results, including a 3rd position result in the recent ICML black box learning challenge as well as competitive results for signal peptide prediction, an important problem in bioinformatics.
Cite
@article{arxiv.1312.5869,
title = {Principled Non-Linear Feature Selection},
author = {Dimitrios Athanasakis and John Shawe-Taylor and Delmiro Fernandez-Reyes},
journal= {arXiv preprint arXiv:1312.5869},
year = {2014}
}
Comments
arXiv admin note: substantial text overlap with arXiv:1311.5636