Comparing Data Assimilation and Likelihood-Based Inference on Latent State Estimation in Agent-Based Models

2025-09-22

AI Summary

This study addresses the estimation of latent states in agent-based models (ABMs) from observable time series driven by partial microstate dynamics. Method: We conduct the first methodological comparison within the ABM framework between data assimilation (DA) and likelihood-based Bayesian inference (LBI), using the bounded-confidence opinion dynamics model as a benchmark. Contribution/Results: LBI significantly outperforms DA in recovering individual-level latent states—such as agent opinions—and exhibits greater robustness to model misspecification. In contrast, DA achieves comparable accuracy in forecasting aggregate population-level statistics and demonstrates superior real-time adaptability. Our analysis delineates the complementary applicability domains of these approaches: LBI excels at microscale inference, whereas DA is better suited for macroscale dynamic calibration. These findings provide an actionable, empirically grounded guideline for selecting appropriate inference methodologies in ABM-driven empirical modeling.

Technology Category

• Intelligent Robots: State Estimation
• Reasoning under Uncertainty: Relational Probabilistic Models
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success

Abstract

In this paper, we present the first systematic comparison of Data Assimilation (DA) and Likelihood-Based Inference (LBI) in the context of Agent-Based Models (ABMs). These models generate observable time series driven by evolving, partially-latent microstates. Latent states need to be estimated to align simulations with real-world data -- a task traditionally addressed by DA, especially in continuous and equation-based models such as those used in weather forecasting. However, the nature of ABMs poses challenges for standard DA methods. Solving such issues requires adaptation of previous DA techniques, or ad-hoc alternatives such as LBI. DA approximates the likelihood in a model-agnostic way, making it broadly applicable but potentially less precise. In contrast, LBI provides more accurate state estimation by directly leveraging the model's likelihood, but at the cost of requiring a hand-crafted, model-specific likelihood function, which may be complex or infeasible to derive. We compare the two methods on the Bounded-Confidence Model, a well-known opinion dynamics ABM, where agents are affected only by others holding sufficiently similar opinions. We find that LBI better recovers latent agent-level opinions, even under model mis-specification, leading to improved individual-level forecasts. At the aggregate level, however, both methods perform comparably, and DA remains competitive across levels of aggregation under certain parameter settings. Our findings suggest that DA is well-suited for aggregate predictions, while LBI is preferable for agent-level inference.

Problem

• Comparing Data Assimilation and Likelihood-Based Inference for latent state estimation in Agent-Based Models
• Estimating partially-latent microstates in ABMs to align simulations with real-world observational data
• Addressing challenges in latent state estimation for opinion dynamics models like Bounded-Confidence Model

Innovation

• Data Assimilation approximates likelihood model-agnostically
• Likelihood-Based Inference uses model-specific likelihood functions
• Comparison performed on Bounded-Confidence opinion dynamics model