Statistical Physics of Agent-Based Modeling for Social Systems: (11 Evolutionary Economics and Social Complexity Science)

This is the first book to provide useful concepts and tools for dealing with agent-based modeling of social systems from the point of view of statistical physics. The statistical physics approach links social scientists with physicists and computer scientists through correlations in empirical data sets. Especially important is that readers will easily be able to recognize that spin systems including the Ising model and its variants can be applied to various social and economic systems as a minimal but universal model to determine the macroscopic properties of an artificial society. It has also been discovered that various empirical data can be built into the model through external fields acting on each spin or interaction between spins. In this sense, the model is regarded as random spin systems including so-called spin glasses. Drastic changes such as bubbles, crashes, or breakdowns in society due to interacting agents are well understood as phase transitions in the literature of physics. This book presents various examples of such remarkable phenomena and provides a useful guide for readers to utilize statistical physics modeling and analysis by applying them to topics such as financial markets, labor markets, housing markets, and social problems. Each interacting agent as a minimum gradient of artificial society can be regarded as a spin, namely, a tiny magnet on an atomic-scale length. By carefully solving the exercises with data analyses in this well-organized book, readers who have no background in physics can easily come to understand the statistical physics approach. For that reason, this book is highly recommended to researchers who seek to learn unconventional ways of thinking about social and economic systems.

Table of Contents:
Ch1. Introduction (1 Age of big data science 2 Conventional statistical and computer scientific tools and their limitation 3 What's new in statistical physics approach).- Ch2. An example: financial market (1 Up-down of market price and stochastic process 2 Stylized fact 3 Correlation between stocks during crises 4 Visualization of correlations).- Ch3. Ising model description (1 Buy and Sell by Ising spin2 Thermodynamics of Ising model 3 Order-disorder phase transition as collective behavior 4 A use of Ising model to describe financial market 5 Bubble and crush as a phase transition) .- Ch4. Inverse problem of Ising model (1 Inference of interaction between agents 2 Data science and inverse Ising problem 3 A relation between inverse Ising problem and network science 4 Some application to financial market).- Ch5. The Potts model: A variant of Ising model (1 Multivariate decision makings 2 Phase transition in the Potts model 3 Application to social problem: labor market 4 Quantifying mismatch between job-seekers and companies 5 Phase transition in labor market.- Ch6. Spatio-temporal pattern formation (1 Housing market in two-dimension 2 Does people purchase the suitable place or their 'neighbour'? 3 Spatial-correlation in rent distribution 4 A relationship between the model and spatial economy).- Ch7. Other examples (1 Social network service and big data 2 Zapping behavior of agent and television commercial market 3 Future direction) .- Ch8. Summary.