Learning Individual Interactions from Population Dynamics with Discrete-Event Simulation Model

Yan Shen,Fan Yang,Mingchen Gao,Wen Dong

2024-05-15

Abstract:The abundance of data affords researchers to pursue more powerful computational tools to learn the dynamics of complex system, such as neural networks, engineered systems and social networks. Traditional machine learning approaches capture complex system dynamics either with dynamic Bayesian networks and state space models, which is hard to scale because it is non-trivial to prescribe the dynamics with a sparse graph or a system of differential equations; or a deep neural networks, where the distributed representation of the learned dynamics is hard to interpret. In this paper, we will explore the possibility of learning a discrete-event simulation representation of complex system dynamics assuming multivariate normal distribution of the state variables, based on the observation that many complex system dynamics can be decomposed into a sequence of local interactions, which individually change the system state only minimally but in sequence generate complex and diverse dynamics. Our results show that the algorithm can data-efficiently capture complex network dynamics in several fields with meaningful events.

Machine Learning

What problem does this paper attempt to address?

The paper aims to address the problem of learning interactions between individuals from the group dynamics of complex systems. Specifically, the authors propose a method based on Discrete-Event Simulation (DES) to represent the dynamics of complex systems. This method assumes state variables with a multivariate normal distribution and observes that many complex system dynamics can be decomposed into a series of local interactions. These interactions individually have a small impact on the system state but sequentially generate complex dynamic changes. The authors use a generalized expectation-maximization algorithm to jointly learn a set of local events and their rates (the number of occurrences per unit time), as well as a coefficient matrix that mixes event counts into changes in state variables. The proposed discrete-event simulator not only demonstrates high-level machine learning performance in predicting unstudied dynamics but also provides useful insights into why the trained model behaves in a specific way and offers an intuitive interface to correct any errors produced by the model. In summary, the goal of the paper is to capture the micro-interactions of complex systems through discrete-event modeling in a data-driven context and to explain the macro-dynamics through these interactions. This approach is applicable to various scenarios such as neural networks, engineering systems, and social networks.

Learning Individual Interactions from Population Dynamics with Discrete-Event Simulation Model

Optimizing Complex Interaction Dynamics in Critical Infrastructure with a Stochastic Kinetic Model.

Integrating Simulation and Signal Processing in Tracking Complex Social Systems.

Integrating Simulation and Signal Processing with Stochastic Social Kinetic Model.

Learning interpretable dynamics of stochastic complex systems from experimental data

Modeling and Simulation of Large-Scale Social Networks Using Parallel Discrete Event Simulation

Learning continuous network emerging dynamics from scarce observations via data-adaptive stochastic processes

Data-driven Discovery of Emergent Behaviors in Collective Dynamics

Expectation Propagation with Stochastic Kinetic Model in Complex Interaction Systems.

Understanding complex crowd dynamics with generative neural simulators

Learning to Simulate Crowd Trajectories with Graph Networks.

A Temporally Heterogeneous Survival Framework with Application to Social Behavior Dynamics

Learning Locally Interacting Discrete Dynamical Systems: Towards Data-Efficient and Scalable Prediction

Modeling information diffusion dynamics over social networks

Decomposed Linear Dynamical Systems (dLDS) for learning the latent components of neural dynamics

Data-driven modeling of interrelated dynamical systems

Towards complex dynamic physics system simulation with graph neural ordinary equations

SUPE-Net: an Efficient Parallel Simulation Environment for Large-Scale Networked Social Dynamics.

Recurrent Switching Dynamical Systems Models for Multiple Interacting Neural Populations