Reconstructing network activity from sampled network data using archetypal analysis

A method for reconstructing network activity, comprising:

• monitoring network activity of a plurality of network entities;
• obtaining a first set of features for each network entity in the plurality of network entities based on the monitoring;
• determining a number of vertices to describe the first sets of features in a multidimensional space;
• assigning a usage pattern to the vertices;
• and
• obtaining a second set of features for a particular network entity that is not included in the plurality of network entities, and to represent the particular network entity based at least on the usage patterns and the second set of features for the particular network entity.

A computer-readable memory device having program code recorded thereon that when executed by at least one processor causes the at least one processor to perform a method comprising:

• monitoring network activity of a plurality of network entities;
• obtaining a first set of features for each network entity in the plurality of network entities based on the monitoring;
• determining a number of vertices to describe the first sets of features in a multidimensional space;
• assigning a usage pattern to the vertices;
• and
• obtaining a second set of features for a particular network entity that is not included in the plurality of network entities, and to represent the particular network entity based at least on the usage patterns and the second set of features for the particular network entity.