A tool for filtering information in complex systems

摘要: We introduce a technique to filter out complex data sets by extracting subgraph of representative links. Such filtering can be tuned up any desired level controlling the genus resulting graph. show that this is especially suitable for correlation-based graphs, giving filtered graphs preserve hierarchical organization minimum spanning tree but containing larger amount information in their internal structure. In particular case planar (genus equal 0), triangular loops and four-element cliques are formed. The application procedure 100 stocks U.S. equity markets shows such have important significant relationships with market structure properties.