The New Multilayer Ensemble Classifier for Verifying Users Based on Keystroke Dynamics

In this work we proposed the new multilayer ensemble classifier which can be applied in many domains, especially in the biometric systems. Proposed classifier works on database which comprises data from keystroke dynamics. Such kind of data allows us to recognize computer users who use password. It is a typical case among the users every day work. Obtained results confirm that proposed multilayer ensemble classifier gives the high security level. For this reason our method can be used to protect computer resources against forgers and imposters.