On selecting interacting features from high-dimensional data

Authors:
Peter Hall;Jing-Hao Xue
Affiliations:
Department of Mathematics and Statistics, The University of Melbourne, VIC 3010, Australia;Department of Statistical Science, University College London, London WC1E 6BT, UK
Venue:
Computational Statistics & Data Analysis
Year:
2014

Citing 7
Cited 1

Correlation-based Feature Selection for Discrete and Numeric Class Machine Learning

ICML '00 Proceedings of the Seventeenth International Conference on Machine Learning
An introduction to variable and feature selection

The Journal of Machine Learning Research
Systematic benchmarking of microarray data classification: assessing the role of non-linearity and dimensionality reduction

Bioinformatics
A review of feature selection techniques in bioinformatics

Bioinformatics
Performance of feature-selection methods in the classification of high-dimension data

Pattern Recognition
Ultrahigh Dimensional Feature Selection: Beyond The Linear Model

The Journal of Machine Learning Research
Knowledge discovery approach to automated cardiac SPECT diagnosis

Artificial Intelligence in Medicine

Estimating mutual information for feature selection in the presence of label noise

Computational Statistics & Data Analysis

Quantified Score

Hi-index	0.03

Visualization

Abstract

For high-dimensional data, most feature-selection methods, such as SIS and the lasso, involve ranking and selecting features individually. These methods do not require many computational resources, but they ignore feature interactions. A simple recursive approach, which, without requiring many more computational resources, also allows identification of interactions, is investigated. This approach can lead to substantial improvements in the performance of classifiers, and can provide insight into the way in which features work together in a given population. It also enjoys attractive statistical properties.