A novel multivariate feature selection methodology using the gamma-metric showed good predictive performance for classification tasks, including atrial fibrillation detection, though it struggled to exclude non-informative features in highly correlated or large datasets.
Abstract Background The γ -metric value is generally used as the importance score of a feature (or a set of features) in a classification context. This study aimed to go further by creating a new methodology for multivariate feature selection for classification, whereby the γ -metric is associated with a specific search direction (and therefore a specific stopping criterion). As three search directions are used, we effectively created three distinct methods. Methods We assessed the performance of our new methodology through a simulation study, comparing them against more conventional methods. Classification performance indicators, number of selected features, stability and execution time were used to evaluate the performance of the methods. We also evaluated how well the proposed methodology selected relevant features for the detection of atrial fibrillation, which is a cardiac arrhythmia. Results We found that in the simulation study as well as the detection of AF task, our methods were able to select informative features and maintain a good level of predictive performance; however in a case of strong correlation and large datasets, the γ -metric based methods were less efficient to exclude non-informative features. Conclusions Results highlighted a good combination of both the forward search direction and the γ -metric as an evaluation function. However, using the backward search direction, the feature selection algorithm could fall into a local optima and can be improved.
Ngo et al. (Thu,) studied this question.