Selecting features with genetic algorithm in handwritten digit recognition
Weiquan Liu, Minghui Wang, Yixin Zhong · 2002
In this paper, a feature selection method is described. Ln a pattern recognition system, a large number of features usually make the realization of eflcient class$er df7cult. The redundancy within features is also unavoidable. The method proposed in this pigper selects features from the existing feature set according to the mutual information(iMIl measuremervt between classes and features. Genelic Algorithm(GA) is used to select the most informative feature subset. Based on the experiment results if handwritten digits recognition, This method can reduce the number of features needed in the recognition process without impairing the performance of class(j7er significantly.