Adaptive techniques for intra-user variability in keystroke dynamics

Conventional machine learning algorithms based on keystroke dynamics build a classifier from labeled data in one or more sessions but assume that the dataset at the time of verification exhibits the same distribution. A user's typing characteristics may gradually change over time and space. Therefore, a traditional classifier may perform poorly on another dataset that is acquired under different environmental conditions. In this paper, we investigate the applicability of transfer learning to update a classifier according to the changing environmental conditions with minimum amount of re-training. We show that by using adaptive techniques, it is possible to identify an individual at a different time by acquiring only a few samples from another session, and at the same time obtain up to 13% higher accuracy. We make a comparative analysis among the proposed algorithms and conclude that adaptive classifiers exhibit a higher start by a good approximation and perform better than the classifier trained from start-over.

[1]  Shambhu J. Upadhyaya,et al.  Enhanced recognition of keystroke dynamics using Gaussian mixture models , 2015, MILCOM 2015 - 2015 IEEE Military Communications Conference.

[2]  Stefan Rüping,et al.  Incremental Learning with Support Vector Machines , 2001, ICDM.

[3]  Lee Luan Ling,et al.  User authentication through typing biometrics features , 2005, IEEE Transactions on Signal Processing.

[4]  Philip S. Yu,et al.  Mining concept-drifting data streams using ensemble classifiers , 2003, KDD '03.

[5]  Christophe Rosenberger,et al.  Analysis of template update strategies for keystroke dynamics , 2011, 2011 IEEE Workshop on Computational Intelligence in Biometrics and Identity Management (CIBIM).

[6]  Arun Ross,et al.  An introduction to biometric recognition , 2004, IEEE Transactions on Circuits and Systems for Video Technology.

[7]  Sungzoon Cho,et al.  Continual Retraining of Keystroke Dynamics Based Authenticator , 2007, ICB.

[8]  William-Chandra Tjhi,et al.  Dual Fuzzy-Possibilistic Co-clustering for Document Categorization , 2007 .

[9]  Michael K. Reiter,et al.  Password hardening based on keystroke dynamics , 1999, CCS '99.

[10]  Arun Ross,et al.  Biometric template selection and update: a case study in fingerprints , 2004, Pattern Recognit..

[11]  Roy A. Maxion,et al.  Comparing anomaly-detection algorithms for keystroke dynamics , 2009, 2009 IEEE/IFIP International Conference on Dependable Systems & Networks.

[12]  Andrew Zisserman,et al.  Tabula rasa: Model transfer for object category detection , 2011, 2011 International Conference on Computer Vision.

[13]  Danoush Hosseinzadeh,et al.  Gaussian Mixture Modeling of Keystroke Patterns for Biometric Applications , 2008, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[14]  Andrey Makrushin,et al.  Automatic Template Update Strategies for Biometrics , 2007 .

[15]  Peter Stone,et al.  Transfer Learning for Reinforcement Learning Domains: A Survey , 2009, J. Mach. Learn. Res..

[16]  Rong Yan,et al.  Adapting SVM Classifiers to Data with Shifted Distributions , 2007, Seventh IEEE International Conference on Data Mining Workshops (ICDMW 2007).

[17]  Gian Luca Marcialis,et al.  Adaptive Biometric Systems That Can Improve with Use , 2008 .

[18]  Gian Luca Marcialis,et al.  Template Update Methods in Adaptive Biometric Systems: A Critical Review , 2009, ICB.

[19]  Zhengrong Liang,et al.  ROC operating point selection for classification of imbalanced data with application to computer-aided polyp detection in CT colonography , 2013, International Journal of Computer Assisted Radiology and Surgery.