aiqbench analysis
This research introduces PosEDO, a novel framework for online calibration of black-box social simulators using posterior-assisted evolutionary dynamic optimization. By leveraging observation-conditioned parameter-space signals, the method effectively detects regime shifts in nonstationary economic and financial models without requiring additional simulator calls. The approach advances machine learning techniques for adapting to complex, changing environments where traditional fitness-based adaptation fails.
Online Regime-aware Calibration for Black-box Social Simulators via Posterior-assisted Evolutionary Dynamic Optimization
arXiv:2601.19481v2 Announce Type: replace-cross Abstract: Evolutionary dynamic optimization (EDO) commonly assumes that environmental changes can be detected from fitness variations and handled through random re-initialization, historical solutions, or learned transition patterns. Online calibration of black-box simulators introduces a different setting, where the dynamic objective is induced by sequential observations and a changing calibration window, rather than being controlled by explicit variables. Fitness variations therefore cannot be directly attributed to regime changes, while the unknown relationship between successive regimes limits conventional adaptation. We formulate this setting as an observation-driven dynamic optimization problem and propose PosEDO, which augments fitness-based EDO with an observation-conditioned parameter-space signal. PosEDO learns this signal online as a posterior distribution over simulator parameters from parameter-trajectory pairs generated during evolutionary evaluation, using posterior shifts for change detection and posterior samples for population adaptation. The new evaluation records are further utilized for online posterior updating without additional simulator calls. Experiments on nonstationary economic and financial simulators show that PosEDO improves calibration accuracy, optimization performance, and change-detection quality over representative EDO baselines.
Abstract:Evolutionary dynamic optimization (EDO) commonly assumes that environmental changes can be detected from fitness variations and handled through random re-initialization, historical solutions, or learned transition patterns. Online calibration of black-box simulators introduces a different setting, where the dynamic objective is induced by sequential observations and a changing calibration window, rather than being controlled by explicit variables. Fitness variations therefore cannot be directly attributed to regime changes, while the unknown relationship between successive regimes limits conventional adaptation. We formulate this setting as an observation-driven dynamic optimization problem and propose PosEDO, which augments fitness-based EDO with an observation-conditioned parameter-space signal. PosEDO learns this signal online as a posterior distribution over simulator parameters from parameter-trajectory pairs generated during evolutionary evaluation, using posterior shifts for change detection and posterior samples for population adaptation. The new evaluation records are further utilized for online posterior updating without additional simulator calls. Experiments on nonstationary economic and financial simulators show that PosEDO improves calibration accuracy, optimization performance, and change-detection quality over representative EDO baselines.
Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG)
arXiv:2601.19481 [cs.NE]
(or arXiv:2601.19481v2 [cs.NE] for this version)
https://doi.org/10.48550/arXiv.2601.19481 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 263 words
Related coverage
MA-SBI: Misspecification-Aware Simulation-Based Inference via Side-Channel Guidance
arXiv · Jun 16, 2026
Uncertainty and Explainability in Deep Rough Volatility: A Neural Information-Theoretic Posterior Approach
arXiv · Sep 28, 2026
Encoded Coordination on the Open Web
LessWrong · Sep 23, 2026
Testing, not presuming, adequacy: calibrating generative social simulators against emergent network structure
arXiv · Sep 24, 2026