An Efficient Prototype Selection Algorithm Based on Spatial Abstraction

Abstract

Nowadays, machine learning and data mining approaches have been applied in huge volumes of data. In order to deal with this big data, techniques for prototype selection have been applied for reducing the data to a manageable volume and, consequently, for reducing the computational resources that are necessary to apply machine learning approaches. In this paper, we propose an efficient approach for prototype selection called PSSA. It adopts the notion of spatial partition for efficiently splitting the dataset in sets of similar instances. In a second step, the algorithm creates one prototype for each spatial partition and, finally, it selects the prototypes of the densest spatial partitions and that are far from other prototypes that were already selected. The approach was evaluated on 15 well-known datasets used in a classification task, and its performance was compared to those of 6 state-of-the-art algorithms, considering two measures: accuracy and reduction. All the obtained results show that, in general, the proposed approach provides a good trade-off between accuracy and reduction, with a significantly lower running time, when compared to other approaches.