Self-organizing neural grove and its applications

Recently, multiple classifier systems (MCS) have been used for practical applications to improve classification accuracy. Self-generating neural networks (SGNN) are one of the suitable base-classifiers for MCS because of their simple setting and fast learning. However, the computation cost of the NICS increases in proportion to the number of SGNN. In this paper, we propose a novel pruning method for efficient classification and we call this model as self-organizing neural grove (SONG). Experiments have been conducted to compare the pruned NICS with an unpruned NICS, the MCS based on C4.5, and k-nearest neighbor method. The results show that the pruned NICS can improve its classification accuracy as well as reducing the computation cost.