Estimating topological properties of weighted networks from limited information

Abstract

A fundamental problem in studying and modeling economic and financial systems is represented by privacy issues, which put severe limitations on the amount of accessible information. Here we introduce a novel, highly nontrivial method to reconstruct the structural properties of complex weighted networks of this kind using only partial information: the total number of nodes and links, and the values of the strength for all nodes. The latter are used as fitness to estimate the unknown node degrees through a standard configuration model. Then, these estimated degrees and the strengths are used to calibrate an enhanced configuration model in order to generate ensembles of networks intended to represent the real system. The method, which is tested on real economic and financial networks, while drastically reducing the amount of information needed to infer network properties, turns out to be remarkably effective$-$thus representing a valuable tool for gaining insights on privacy-protected socioeconomic systems.

References in corpus (3)

• Hierarchical structure and the prediction of missing links in networks
• A fitness model for the Italian Interbank Money Market
• Generalized Bose-Fermi statistics and structural correlations in weighted networks

Cited by in corpus (28)

• The Statistical Physics of Real-World Networks
• The Physics of Financial Networks
• DebtRank: A microscopic foundation for shock propagation
• Reconstruction methods for networks: the case of economic and financial systems
• Distress propagation in complex networks: the case of non-linear DebtRank
• Fast and scalable likelihood maximization for Exponential Random Graph Models with local constraints
• The Structure of Online Social Networks Mirror Those in the Offline World
• The Internet of People: A human and data-centric paradigm for the Next Generation Internet
• Network structure, metadata and the prediction of missing nodes and annotations
• Enhanced capital-asset pricing model for the reconstruction of bipartite financial networks
• Reconstructing networks
• The Grand Canonical ensemble of weighted networks
• Configuration model for correlation matrices preserving the node strength
• Reconstructing firm-level interactions: the Dutch input-output network
• Entangling credit and funding shocks in interbank markets
• A faster horse on a safer trail: generalized inference for the efficient reconstruction of weighted networks
• Systemic liquidity contagion in the European interbank market
• Network reconstruction via density sampling
• Irreducible network backbones: unbiased graph filtering via maximum entropy
• Social dynamics of financial networks
• Entropy-based approach to missing-links prediction
• Controlling systemic risk - network structures that minimize it and node properties to calculate it
• Information Recovery In Behavioral Networks
• A maximum entropy network reconstruction of macroeconomic models
• Reconstruction of Interbank Network using Ridge Entropy Maximization Model
• Critical density for network reconstruction
• Reconstruction of multiplex networks with correlated layers
• Interbank network reconstruction enforcing density and reciprocity