In this paper, we propose a novel continuous authentication system forsmartphone users. The proposed system entirely relies on unlabeled phonemovement patterns collected through smartphone accelerometer. The data wascollected in a completely unconstrained environment over five to twelve days.The contexts of phone usage were identified using k-means clustering. Multipleprofiles, one for each context, were created for every user. Five machinelearning algorithms were employed for classification of genuine and impostors.The performance of the system was evaluated over a diverse population of 57users. The mean equal error rates achieved by Logistic Regression, NeuralNetwork, kNN, SVM, and Random Forest were 13.7%, 13.5%, 12.1%, 10.7%, and 5.6%respectively. A series of statistical tests were conducted to compare theperformance of the classifiers. The suitability of the proposed system fordifferent types of users was also investigated using the failure to enrollpolicy.
展开▼
