To diminish the behavioral variability of mouse dynamics, the machine learning algorithm was proposed. Mouse dynamics is the process of identifying the user based on their mouse operating behavior. The dataset includes co-ordinates values, time stamp value and mouse operation. From this dataset, the schematic features, holistic features and motor-skill features like average speed, average distance, mean, standard deviation and mouse silence ratio, velocity, slope angle, curvature were extracted to obtain feature vector. The obtained feature vector can be applied to the dimensionality reduction based approach, diffusion map to reduce the dimension of the feature vector that compared with ISOMAP (Isometric Feature Mapping). Without dimensionality reduction based method the classification process was difficult. The machine learning algorithm i.e.) hop field network to be used to identify whether the given input sample was authenticated user (or) unauthenticated.