Maximum likelihood: extracting unbiased information from complex networks

Abstract

The choice of free parameters in network models is subjective, since it depends on what topological properties are being monitored. However, we show that the Maximum Likelihood (ML) principle indicates a unique, statistically rigorous parameter choice, associated to a well defined topological feature. We then find that, if the ML condition is incompatible with the built-in parameter choice, network models turn out to be intrinsically ill-defined or biased. To overcome this problem, we construct a class of safely unbiased models. We also propose an extension of these results that leads to the fascinating possibility to extract, only from topological data, the `hidden variables' underlying network organization, making them `no more hidden'. We test our method on the World Trade Web data, where we recover the empirical Gross Domestic Product using only topological information.

Final version accepted for publication

References in corpus (15)

• The statistical mechanics of networks
• Mixture models and exploratory analysis in networks
• Class of correlated random networks with hidden variables
• Fitness-dependent topological properties of the World Trade Web
• A General Formalism for Inhomogeneous Random Graphs
• The scale-free topology of market investments
• Local graph alignment and motif search in biological networks
• The origin of degree correlations in the Internet and other networks
• Statistical ensemble of scale-free random graphs
• Correlated random networks
• A system of mobile agents to model social networks
• Multi-species grandcanonical models for networks with reciprocity
• Inversion method for content-based networks
• Phase transitions in social networks
• Fluctuation-dissipation relations for complex networks

Cited by in corpus (99)

• The Statistical Physics of Real-World Networks
• Entropy measures for complex networks: Toward an information theory of complex topologies
• The entropy of network ensembles
• Hierarchical Block Structures and High-resolution Model Selection in Large Networks
• Parsimonious module inference in large networks
• Efficient Monte Carlo and greedy heuristic for the inference of stochastic block models
• Analytical maximum-likelihood method to detect patterns in real networks
• Early-warning signals of topological collapse in interbank networks
• Randomizing bipartite networks: the case of the World Trade Web
• Community detection for correlation matrices
• Randomizing world trade. I. A binary network analysis
• Randomizing world trade. II. A weighted network analysis
• Inferring monopartite projections of bipartite networks: an entropy-based approach
• Enhanced reconstruction of weighted networks from strengths and degrees
• The role of bot squads in the political propaganda on Twitter
• Self-organized network evolution coupled to extremal dynamics
• Breaking of ensemble equivalence in networks
• Reconstruction methods for networks: the case of economic and financial systems
• Clustering implies geometry in networks
• Reconstructing the world trade multiplex: the role of intensive and extensive biases
• The weighted random graph model
• Null Models of Economic Networks: The Case of the World Trade Web
• Breaking the spell of nestedness
• Flow of online misinformation during the peak of the COVID-19 pandemic in Italy
• Bitcoin Transaction Networks: an overview of recent results
• Fast and scalable likelihood maximization for Exponential Random Graph Models with local constraints
• Complex Networks and Symmetry I: A Review
• Percolation in self-similar networks
• Enhanced Gravity Model of trade: reconciling macroeconomic and network models
• Network comparison and the within-ensemble graph distance
• Extracting significant signal of news consumption from social networks: the case of Twitter in Italian political elections
• Brexit and bots: characterizing the behaviour of automated accounts on Twitter during the UK election
• Entropy distribution and condensation in random networks with a given degree distribution
• Clustering of random scale-free networks
• A GDP-driven model for the binary and weighted structure of the International Trade Network
• Firms' Challenges and Social Responsibilities during Covid-19: a Twitter Analysis
• Triadic motifs and dyadic self-organization in the World Trade Network
• Reconstructing firm-level interactions: the Dutch input-output network
• Multiplexity versus correlation: the role of local constraints in real multiplexes
• Complex Networks and Symmetry II: Reciprocity and Evolution of World Trade
• A faster horse on a safer trail: generalized inference for the efficient reconstruction of weighted networks
• Testing structural balance theories in heterogeneous signed networks
• Multiplexity and multireciprocity in directed multiplexes
• Sparse Maximum-Entropy Random Graphs with a Given Power-Law Degree Distribution
• Grand canonical validation of the bipartite International Trade Network
• Hidden Variables in Bipartite Networks
• Ensemble nonequivalence in random graphs with modular structure
• Meta-validation of bipartite network projections
• Statistical Mechanics of Multi-Edge Networks
• The ambiguity of nestedness under soft and hard constraints
• Irreducible network backbones: unbiased graph filtering via maximum entropy
• Sustainable Development Goals as unifying narratives in large UK firms' Twitter discussions
• Bow-Tie Structures of Twitter Discursive Communities
• Stationarity, non-stationarity and early warning signals in economic networks
• Entropy-based approach to missing-links prediction
• Spatial effects in real networks: measures, null models, and applications
• The role of distances in the World Trade Web
• Entropy-based detection of Twitter echo chambers
• Entropy-based randomisation of rating networks
• Random Simplicial Complexes: Models and Phenomena
• Weighted hypersoft configuration model
• The role of adjacency matrix degeneration in maximum entropy weighted network models
• From Ecology to Finance (and Back?): Recent Advancements in the Analysis of Bipartite Networks
• Covariance structure behind breaking of ensemble equivalence in random graphs
• Maximum Entropy approach to multivariate time series randomization
• Colombian export capabilities: building the firms-products network
• Bibliometric and social network analysis on the use of satellite imagery in agriculture: an entropy-based approach
• Thermodynamics of network model fitting with spectral entropies
• A Statistical Test of Walrasian Equilibrium by Means of Complex Networks Theory
• Entropy of labeled versus unlabeled networks
• Self-Organization and Complex Networks
• Binary versus non-binary information in real time series: empirical results and maximum-entropy matrix models
• Hamiltonian Dynamics of Preferential Attachment
• Sparse power-law network model for reliable statistical predictions based on sampled data
• Entropy-based models to randomize real-world hypergraphs
• Interactions between financial and environmental networks in OECD countries
• The double role of GDP in shaping the structure of the International Trade Network
• Ensemble nonequivalence and Bose-Einstein condensation in weighted networks
• Online disinformation in the 2020 U.S. Election: swing vs. safe states
• Statistical physics of exchangeable sparse simple networks, multiplex networks and simplicial complexes
• Inferring comparative advantage via entropy maximization
• Reconstructing topological properties of complex networks using the fitness model
• A maximum entropy network reconstruction of macroeconomic models
• Stochastic Evolving Differential Games Toward a Systems Theory of Behavioral Social Dynamics
• Economic networks in and out of equilibrium
• Correspondence between temporal correlations in time series, inverse problems, and the Spherical Model
• Patterns of link reciprocity in directed, signed networks
• Temporal networks with node-specific memory: unbiased inference of transition probabilities, relaxation times and structural breaks
• Linking Through Time: Memory-Enhanced Community Discovery in Temporal Networks
• Learn your entropy from informative data: an axiom ensuring the consistent identification of generalized entropies
• Critical density for network reconstruction
• A null model for Dunbar's circles
• Statistically validated projection of bipartite signed networks
• Maximum entropy modeling of Optimal Transport: the sub-optimality regime and the transition from dense to sparse networks
• Reconstruction of multiplex networks with correlated layers
• Assessing imbalance in signed brain networks
• Leveraging Content Producer Networks and User Perception to Detect Online Discursive Communities
• Italian Twitter semantic network during the Covid-19 epidemic
• Commodity-specific triads in the Dutch inter-industry production network