Network Inference from Co-Occurrences

The study of networked systems is an emerging field, impacting almost every area of engineering and science, including the important domains of communication systems, biology, sociology, and cognitive science. The recovery of network structure from experimental data is a basic and fundamental problem. Unfortunately, experimental data often do not directly reveal network structure due to inherent measurement limitations such as imprecision in timing or other observation mechanisms. This paper considers the following problem. Suppose a number of transmissions are made between a collection of senders and receivers. We observe the subset of network elements (e.g., communication links, genes, actors, neuron colonies) which carry each transmission, but the order in which these elements appear in the transmission paths is not observable. Mathematically, the network structure can be described by a graph whose vertices are the communicating elements, senders and receivers. Each transmission is...

• Type: Journal
• Year: 2006
• Where: CORR
• Authors: Michael Rabbat, Mário A. T. Figueiredo, Robert Nowak
• Keywords: CORR 2006, Education, Feasible Graphs, Inherent Measurement Limitations, Network Structure