A Preselection-Based Fast Support Vector Machine Learning for Large-Scale Pattern Sets using Compressed Relative Neighborhood Graph

We propose a pre-selection method for training support vector machines (SVM) with a largescale dataset. Specifically, the proposed method selects patterns around the class boundary and the selected data is fed to train an SVM. For the selection, that is, searching for boundary patterns, we utilize a compressed representation of relative neighborhood graph (Clustered-RNG). A Clustered-RNG is a network of neighboring patterns which have a different class label and thus, we can find boundary patterns between different classes. Through large-scale handwritten digit pattern recognition experiments, we show that the proposed pre-selection method accelerates SVM training process 10 times faster without degrading recognition accuracy