Dynamics on and of complex networks III : machine learning and statistical physics approaches

Intro; Preface; List of Reviewers (Alphabetically Ordered by Last Names); Contents; Part I Network Structure; An Empirical Study of the Effect of Noise Modelson Centrality Metrics; 1 Introduction; 2 Experimental Setup; 2.1 Test Suite of Networks; 2.2 Centrality Metrics; 2.3 Methodology; 3 Empirical Results; 3.1 Edge Addition; 3.2 Edge Deletion; 3.3 Edge Swap; 3.4 Edge XOR; 3.5 Summary of the Results; 4 Related Research; 5 Conclusion and Future Work; References; Emergence and Evolution of Hierarchical Structurein Complex Systems; 1 Introduction; 2 Lexis Background; 2.1 Lexis-DAG 2.2 The Lexis Optimization Problem2.3 Path Centrality and the Core of a Lexis-DAG; 2.4 Hourglass Score; 3 Evo-Lexis Framework and Metrics; 3.1 Incremental Design Algorithm; 3.2 Target Generation Models; 3.2.1 MRS Model; 3.2.2 MS Model; 3.2.3 M Model; 3.2.4 RND Model; 3.3 Key Metrics; 3.3.1 Cost Metrics; 3.3.2 Topological Metrics; 3.3.3 Target Diversity Metric; 4 Computational Results; 4.1 Parameter Values and Evolutionary Iteration; 4.2 Results; 4.2.1 Emergence of Low-Cost Hierarchies Due to Tinkering/Mutation and Selection 4.2.2 Low-Cost Design Resulting in Deeper Hierarchies and Reuse of More Complex Modules4.2.3 The Recombination Mechanism Creates Target Diversity; 4.2.4 Reuse of Complex Modules in the Core Set by Strong Selection; 4.2.5 Emergence of Hourglass Architecture Due to the Heavy Reuse of Complex Intermediate Modules in Models with Strong Selection; 4.2.6 Stability of the Core Set Due to Selection; 4.2.7 Fragility Caused by Stronger Selection; 5 Evolvability and the Space of Possible Targets; 6 Major Transitions; 7 Overhead of Incremental Design; 8 Discussion and Prior Work 8.1 Modularity and Hierarchy8.2 Hourglass Architecture; 8.3 Interplay of Design Adaptation and Evolution; 8.4 From Abstract Modeling to Specific Evolving Systems; 9 Conclusion; References; Evaluation of Cascading Infrastructure Failures and Optimal Recovery from a Network Science Perspective; 1 Introduction; 2 Risk and Resiliency; 2.1 Assessing Risk; 2.2 Gaps in the Risk Literature; 2.3 Moving Towards Resilience; 3 Network Science as a Tool; 4 Case Studies; 4.1 Studying Resilience Curves; 4.2 Data; 4.3 Limitations of the Data; 4.4 Network Analysis of IEEE Bus Test Case; 4.5 Network Robustness 4.6 Network Recovery4.7 Universal Resilience Curves [15]; 4.8 Insights and Conclusions; References; Part II Network Dynamics; Automatic Discovery of Families of Network Generative Processes; 1 Introduction; 2 Network Morphogenesis; 2.1 Reconstructing Processes; 2.1.1 Using Micro-Level Processes; 2.1.2 Using Macro-Level Structure; 2.2 Reconstructing Structure; 2.2.1 Using Processes; 2.2.2 Using Structure; 2.3 Combining Both: Evolutionary Models; 3 Symbolic Regression of Network Generators; 4 Families of Network Generators; 4.1 Protocol; 4.2 A Measure of Generator Dissimilarity; 4.3 Two-Dimensional Embedding and Families