Representative-based active learning with max-min distance

Abstract

Active learning has been a hot topic because labeled data are useful, however expensive. Many existing approaches are based on decision trees, Naïve Bayes algorithms, etc. In this paper, we propose a representative-based active learning algorithm with max-min distance. Our algorithm has two techniques interacting with each other. One is the representative-based classification inspired by covering-based neighborhood rough sets. The other is critical instance selection with max-min distance. Experimental results on six UCI datasets indicate that, with the same number of labeled instances, our algorithm is comparable with or better than the ID3, C4.5 and Naïve Bayes algorithms.