Classifier subset selection (CSS) from a large ensemble is an effective way to design multiple classifier systems (MCSs). Given a validation dataset and a selection criterion, the...
Abstract. In adversarial classification tasks like spam filtering, intrusion detection in computer networks and biometric authentication, a pattern recognition system must not only...
The existing methods for offline training of cascade classifiers take a greedy search to optimize individual classifiers in the cascade, leading inefficient overall performance. W...
This paper presents an investigation into the combination of different classifiers for toxicity prediction. These classification methods involved in generating classifiers for comb...
Cross-language document retrieval systems require support by some kind of multilingual thesaurus for semantically indexing documents in different languages. The peculiarities of t...