Ciena Corporation
Network system modeling using nested models combining machine learning and behavioral approaches

Abstract:

Systems and methods include obtaining input data related to a networking system; modeling operation of the networking system via a composite model that includes a plurality of sub-models that collectively form the composite model, wherein at least one sub-model is a behavioral sub-model and at least one sub-model is a machine learning model, wherein various data in the input data is provided to corresponding sub-models, and wherein each sub-model is configured to model one or more components in the networking system; and providing output data based on the modeling.