Feature selection for pattern classification problems
Li Zhang, Gang Sun, Jun Hai Guo · 2004
In pattern recognition feature selection is an important problem which is to choose the smallest subset of features that ideally is necessary and sufficient to describe the target concept. In this paper, a feature selection algorithm based on DB index rules is proposed involving classification capabilities of feature vectors and correlation analysis between two features. The strategy can be used for supervised or unsupervised classification problems and it is evaluated by using three synthetic data sets and a real-word data set.