Equation of state for agents on graphs

TL;DR

This paper develops a statistical thermodynamics framework for modeling agent populations on graphs, deriving equations of state that reveal analogies to classical and quantum gases, and discussing network-specific features.

Contribution

It introduces a novel thermodynamic approach to agent choice models on graphs, deriving equations of state for various network configurations and regimes.

Findings

Equations of state are derived for different graph connectivities.

Analogies to classical and quantum gases are established.

Network-specific features are identified and discussed.

Abstract

Choice models for populations of agents on graphs are studied in terms of statistical thermodynamics. Equations of state are derived and discussed for different connectivity schemes, utility approximations, and temperature and volume regimes. Analogies to ideal classical and quantum gases are found and features specific for network systems are discussed.