Correlation-based Feature Selection

Measures correlation between nominal characteristics so that numerical characteristics are first discretized. CFS is a fully automatic algorithm that does not require the user to specify any thresholds or the number of features to be selected. It operates on the original feature space, meaning that any knowledge induced by a learning algorithm, using features selected by CFS, can be interpreted in terms of the original features, not in terms of a transformed space. This method is also a filter and does not incur high computational cost associated with repeatedly invoking a learning algorithm.