This manuscript presents a supervised machine learning approach in the identification of network attacks on a fingerprint biometric system. To reduce the problem of malicious acts on a biometric system, this manuscript proposes an intrusion detection technique that analyses the fingerprint biometric network traffic for evidence of intrusion. The neural network algorithm that imitates the way a human brain works is used in this study to classify normal traffic and learn the correct traffic pattern on a fingerprint biometric system. The aim of the study is to observe the ability of the neural network in the detection of known and unknown attacks without using a vast amount of training data. The results of the neural network model had a classification rate of 98 %, which translates to a false positive rate of 2%.
Reference:
Mgabile, T, Msiza, IS and Dube, E. Anomaly based intrusion detection for a biometric identification system using neural networks. Planetary Scientific Research Centre, Dubai (UAE), 6-7 October 2012
Mgabile, T., Msiza, I., & Dube, E. (2012). Anomaly based intrusion detection for a biometric identification system using neural networks. Planetary Scientific Research Center (PSRC). http://hdl.handle.net/10204/6196
Mgabile, T, IS Msiza, and E Dube. "Anomaly based intrusion detection for a biometric identification system using neural networks." (2012): http://hdl.handle.net/10204/6196
Mgabile T, Msiza I, Dube E, Anomaly based intrusion detection for a biometric identification system using neural networks; Planetary Scientific Research Center (PSRC); 2012. http://hdl.handle.net/10204/6196 .