Abstract:Random numbers are at the heart of every agent-based model (ABM) of health and disease. By representing each individual in a synthetic population, agent-based models enable detailed analysis of intervention impact and parameter sensitivity. Yet agent-based modeling has a fundamental signal-to-noise problem, in which small changes between simulations cannot be reliably differentiated from stochastic noise resulting from misaligned random number realizations. We introduce a novel methodology that eliminates noise due to misaligned random numbers, a first for agent-based modeling. Our approach enables meaningful individual-level analysis between ABM scenarios because all differences are driven by mechanistic effects rather than random number noise. We demonstrate the benefits of our approach on three disparate examples. Results consistently show reductions in the number of simulations required to achieve a given standard error with levels exceeding 10-fold for some applications.

What problem does this paper attempt to address?

This paper attempts to solve the signal - to - noise ratio problem encountered in comparative analysis in Agent - Based Models (ABMs). Specifically, when using different random number sequences to compare two similar simulation scenarios, the noise caused by the inconsistency of random numbers will mask the real mechanism effects, making it difficult to distinguish small but meaningful differences between the two simulations. This noise problem is particularly prominent when evaluating the impact of interventions, parameter sensitivity analysis, and scenario analysis. ### Main contributions of the paper 1. **Introduction of a new methodology**: - By using Common Random Numbers (CRNs), this method can eliminate the noise caused by inconsistent random numbers, so that the differences between different simulations are completely driven by mechanism effects rather than random noise. - This method for the first time achieves a perfect match between counterfactuals and baseline simulations in Agent - Based Models, significantly improving the reliability and interpretability of simulation results. 2. **Application examples**: - The paper demonstrates the effectiveness of this method through three different application examples: - **Postpartum hemorrhage prevention**: In a model that includes birth and death but no transmission, evaluate the effect of postpartum hemorrhage prevention interventions. - **Vaccination**: In the Susceptible - Infected - Recovered (SIR) transmission model of a closed population and static network, evaluate the effect of vaccination. - **Voluntary Medical Male Circumcision (VMMC)**: In an open population and dynamic transmission network, evaluate the impact of VMMC on the Human Immunodeficiency Virus (HIV). 3. **Performance improvement**: - The results show that using the CRN method can significantly reduce the number of simulations required to reach a given standard error, with a reduction of more than 10 times in some applications. ### Method overview - **Separate pseudo - random number streams**: Assign an independent pseudo - random number generator (PRNG) stream to each decision, ensuring that the same decision uses the same random number seed in different simulations. - **Time - step - dependent PRNG stream jumps**: At each time step, reset each PRNG stream to the initial state and then jump according to the time step, ensuring that each decision stream is in a unique and known state at each time step. - **Slot - based random number allocation**: Assign a "slot" to each agent for indexing the random number drawn from a specific decision PRNG stream at each time step. - **Paired random numbers**: For random numbers that need to act on a pair of individuals, use the paired random number method to ensure that other random numbers do not change even when agents are added or removed. - **Network edge formation**: Propose several network algorithms that maintain CRN consistency, including the Erdős - Rényi dynamic random network, the dynamic disk graph, and the topological embedding method. ### Conclusion Through these innovations, the paper provides an effective method to reduce random noise in Agent - Based Models, enabling researchers to more accurately evaluate the effectiveness of interventions and the sensitivity of parameters. This method is not only applicable to the field of computational epidemiology but can also be widely applied to other research fields that require Agent - Based Models.

Noise-free comparison of stochastic agent-based simulations using common random numbers

Automated and Distributed Statistical Analysis of Economic Agent-Based Models

Using machine learning as a surrogate model for agent-based simulations

Bayesian calibration of stochastic agent based model via random forest

Agent based simulators for epidemic modelling: Simulating larger models using smaller ones

Comparing Stochastic Differential Equations and Agent-Based Modelling and Simulation for Early-Stage Cancer.

Biased versus unbiased numerical methods for stochastic simulations

A Comparison of Economic Agent-Based Model Calibration Methods

Optimizing agent-based transmission models for infectious diseases

A comparison of emulation methods for Approximate Bayesian Computation

Economic evaluations with agent-based modelling: an introduction

An agent-based epidemics simulation to compare and explain screening and vaccination prioritisation strategies

Stochastic Agent-Based Simulations of Social Networks

Calibration of stochastic, agent-based neuron growth models with Approximate Bayesian Computation

CoV-ABM: A stochastic discrete-event agent-based framework to simulate spatiotemporal dynamics of COVID-19

Sensitivity analysis of agent-based models: a new protocol

Private Agent-Based Modeling

The Distance Between: An Algorithmic Approach to Comparing Stochastic Models to Time-Series Data

Black-box Bayesian inference for agent-based models