Maintaining diversity and increasing the accuracy of classification rules through automatic speciation

Multiple species weighted voting (MSWV) is a genetics-based machine learning (GBML) system. MSWV uses two levels of speciation to achieve various objectives. Different species of individuals are by design assigned to each class in the data set. During training, a second level of speciation is achieved when similar individuals are allowed to automatically cluster and form subspecies. In this paper, we are going to show the importance of the automatic speciation in increasing the classification accuracy by maintaining the diversity and increasing the accuracy of the decision rules discovered. Using thirty-six real-world learning tasks we show that MSWV significantly outperforms a number of well known classification algorithms.