Amortized neural networks for agent-based model forecasting

TL;DR

This paper introduces an amortized neural network approach for fast, flexible agent-based model forecasting, capable of generating predictions without re-estimation after initial training.

Contribution

It presents a novel two-step algorithm combining simulation and neural network training for efficient agent-based model forecasting.

Findings

Significantly faster predictions compared to traditional methods

Can generate forecasts for various datasets without re-training

Effective in both unconditional and conditional forecasting scenarios

Abstract

In this paper, we propose a new procedure for unconditional and conditional forecasting in agent-based models. The proposed algorithm is based on the application of amortized neural networks and consists of two steps. The first step simulates artificial datasets from the model. In the second step, a neural network is trained to predict the future values of the variables using the history of observations. The main advantage of the proposed algorithm is its speed. This is due to the fact that, after the training procedure, it can be used to yield predictions for almost any data without additional simulations or the re-estimation of the neural network