Northwestern University researchers Roger Guimera and Marta Sales-Pardo have developed a universal method that can correctly analyze a variety of complex networks. The researchers tested their method on a range of five networks: a karate club, a social network of dolphins, the neural network of a worm, the air transportation network of Eastern Europe, and the metabolic network of Escherichia coli. For each of the five networks, the researchers introduced errors and applied an algorithm to the distorted network. Each time, the algorithm created a new network with the errors separated out, and each new network construction was closer to the original true network.
"The flexibility of our approach, along with its generality and its performance, will make it applicable to many areas where network data reliability is a source of concern," say Guimera and Sales-Pardo. The central idea to the new method is that, even though every network has unique characteristics, they all have nodes that can be put into specific groups, with the nodes connecting to each other based on group membership. The method averages all possible groupings of the nodes and gives each group a weight that reflects its explanatory power.
"There are many ways to map nodes in a network, not just one," says Sales-Pardo. "We consider all the possible ways. By taking the sum of them all, we can identify both missing and spurious connections."
From Northwestern University News Center (IL)
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA