Two-Dimensional Concept Vector Machines Based on an Ionic Interaction Model

Support vector machines (SVMs) have shown state-of-the-art performance for many applications. However, SVMs suffer from high computational complexity for classifying new data points when the number of support vectors is large, since the decision boundary is represented as a linear combination of the support vectors. In this paper, we propose a two-dimensional concept vector machine based on an ionic interaction model. It generates a simple decision boundary by a small number of concept vectors. We compared its performance with SVMs.