Records the keystroke timing data and creates a model. Applies learning algorithms and statistical models to differentiate between a valid user and an intruder

This repository contains the code for the MAEAD method and its implementation

This algorithm exploits the relationships between variables to improve the reconstruction performance of the variational autoencoder (VAE). A correlation score was used as the metric to group the features via a distance-based clustering method. The resulting clusters served as inputs for the Attention-Based VAE.

This engine will be the core of our monitoring mechanism. This engine will use the benefits of machine learning to provide a better solution with dynamic parameters.