Blog: Dartmouth Researchers Develop Computational Tool to Untangle Complex Data

Dartmouth Researchers Develop Computational Tool to Untangle Complex Data
Dartmouth News (12/16/08) Knapp, Susan

Dartmouth College researchers have developed the Partition Decoupling Method (PDM), a mathematical tool that can be used to untangle the underlying structure of time-dependent, interrelated, complex data. "With respect to the equities market, we created a map that illustrated a generalized notion of sector and industry, as well as the interactions between them, reflecting the different levels of capital flow, among and between companies, industries, sectors, and so forth," says Dartmouth professor Daniel Rockmore, who led the development effort. "In fact, it is this idea of flow, be it capital, oxygenated blood, or political orientation, that we are capturing." Capturing flow patterns is critical to understanding the subtle interdependencies found in the different components of complex systems. Analysis of the network of correlations is done using spectral analysis. The analysis is combined with statistical learning tools to uncover regions where the flow circulates more than would be expected at random. The result creates a detailed analysis of the interrelations and provides a wide view of the coarse-scale flow as a whole. Rockmore says PDM uses a different approach than similar programs designed to find how complex systems behave, and because it is not strictly hierarchical, PDM does not constrain interconnectivity.

View Full Article