Description: Based isometric map (ISOMAP) non-linear dimensionality reduction algorithm, we propose a new feature based on the user' s keystrokes the user authentication algorithm with geodesic distance instead of the traditional Euclidean distance between the sample vector as distance measure, the feature vector space in the user' s keystrokes to mine low-Wei Liman embedded manifold, for user identification. 1 with the collected data for 500 Typing Mode experimental test results show that this algorithm outperforms existing similar algorithm, the false rejection rate (FRR) and error through rate (FAR) is 1.65 , respectively, and 0 , lower than existing similar algorithms.
To Search:
File list (Check if you may need any files):
基于流形学习的用户身份认证.pdf