Multiple Classifier Systems

Theoretical issues

Multiple Classifier Systems are a state-of-the-art approach for classifier design. They can overcome several drawbacks of the traditional approach, which is based on designing a single, "monolithic" classification algorithm. They are the methodological core of PRA Lab's research activities.



We exploited the paradigm of Multiple Classifier Systems in nearly all application fields we have addressed so far, including biometric identity recognition, computer and network security, content-based image retrieval, and classification of text, documents and images.