The Multi-Purpose Network

It provides functionalities to model the energy behavior of each node, taking into account the opening/working hours, the level of occupancy of the buildings and the impact of the weather conditions. A dedicated benchmark analysis offers the opportunity to identify efficiency activities and to track their performances.

In Base the definition of site clusters to reach a partitioned view of the network and the consequent energy model, has been evolved to make use of Neural Networks (NN). The algorithms implemented use Decision Trees. Decision trees are non-parametric algorithms that create a statistical function consisting of a multitude of rules applied to nodes with a tree structure that links the energy units features with the target variable through supervised training.