Adaptive Activity Driven Multi‐Level Hierarchical Prediction of Complex Systems Through Profiling and Feedback

Complex systems composed of multi‐level hierarchical systems exhibit complex interacting dependences which affects the performance of the overall system. Although, one solution is to tightly monitor each system activity and feedback to the next step level in order to take appropriate measures and actions to improve the system this solution is costly and too systematic. In addition, a dependency chain exists between low level systems and higher level systems this dependency chains being as long as the number of levels of the system. We propose in this paper an activity driven adaptive hierarchical prediction technique for complex systems which minimizes monitoring and prediction resources requirements and still keep efficient overall systems performance prediction.