Hit Miss Networks with Applications to Instance Selection

In supervised learning, a training set consisting of labeled instances is used by a learning algorithm for generating a model (classifier) that is subsequently employed for deciding the class label of new instances (for generalization). Characteristics of the training set, such as presence of noisy instances and size, influence the learning algorithm and affect generalization performance. This paper introduces a new network-based representation of a training set, called hit miss network (HMN), which provides a compact description of the nearest neighbor relation over pairs of instances from each pair of classes. We show that structural properties of HMN's correspond to properties of training points related to the one nearest neighbor (1-NN) decision rule, such as being border or central point. This motivates us to use HMN's for improving the performance of a 1-NN, classifier by removing instances from the training set (instance selection). We introduce three new HMN-based algorithms for instance selection. HMN-C, which removes instances without affecting accuracy of 1-NN on the original training set, HMN-E, based on a more aggressive storage reduction, and HMN-EI, which applies iteratively HMN-E. Their performance is assessed on 22 data sets with different characteristics, such as input dimension, cardinality, class balance, number of classes, noise content, and presence of redundant variables. Results of experiments on these data sets show that accuracy of 1-NN classifier increases significantly when HMN-EI is applied. Comparison with state-of-the-art editing algorithms for instance selection on these data sets indicates best generalization performance of HMN-EI and no significant difference in storage requirements. In general, these results indicate that HMN's provide a powerful graph-based representation of a training set, which can be successfully applied for performing noise and redundance reduction in instance-based learning.