The Practicality of Normalizing Flow Test-Time Training in Bayesian Inference for Agent-Based Models
By: Junyao Zhang, Jinglai Li, Junqi Tang
Potential Business Impact:
Lets computer models adjust instantly to new information.
Agent-Based Models (ABMs) are gaining great popularity in economics and social science because of their strong flexibility to describe the realistic and heterogeneous decisions and interaction rules between individual agents. In this work, we investigate for the first time the practicality of test-time training (TTT) of deep models such as normalizing flows, in the parameters posterior estimations of ABMs. We propose several practical TTT strategies for fine-tuning the normalizing flow against distribution shifts. Our numerical study demonstrates that TTT schemes are remarkably effective, enabling real-time adjustment of flow-based inference for ABM parameters.
Similar Papers
Amortized Inference of Multi-Modal Posteriors using Likelihood-Weighted Normalizing Flows
Machine Learning (CS)
Finds hidden patterns in complex data faster.
Towards agent-based-model informed neural networks
Machine Learning (CS)
Helps computers understand how groups of things act.
Towards agent-based-model informed neural networks
Machine Learning (CS)
Teaches computers to understand how groups act.