User Behavior Analysis with Machine Learning Techniques in Cloud Computing Architectures

This paper presents the use of machine learning algorithms to analyze the behaviors of users working in a distributed computer environment. The objective consists in discriminating groups of close users. These groups are composed of users with similar behaviors. Event related to the user's behaviors are recorded and transferred to a database. An approach is developed to determine the groups of the users. A non-parametric method of estimating a probability density is used to predict application launches and session openings in an individual way for each user. These algorithms have been implemented and demonstrated their effectiveness within a complete virtualization environment for workstations and applications under real conditions in a hospital.