Agent-Based Model Calibration Using Machine Learning Surrogates

LEM Working Paper

36 Pages Posted: 30 Mar 2017 Last revised: 3 Oct 2017

See all articles by Francesco Lamperti

Francesco Lamperti

Scuola Superiore Sant'Anna di Pisa - Institute of Economics and LEM; Fondazione Eni Enrico Mattei (FEEM)

Andrea Roventini

Scuola Superiore Sant'Anna di Pisa - Laboratory of Economics and Management (LEM); Observatoire Fran├žais des Conjonctures Economiques (OFCE)

Amir Sani

Simudyne Ltd; Scuola Superiore Sant'Anna di Pisa - Laboratory of Economics and Management (LEM)

Date Written: October 3, 2017

Abstract

Efficiently calibrating agent-based models (ABMs) to real data is an open challenge. This paper explicitly tackles parameter space exploration and calibration of ABMs by combining machine-learning and intelligent iterative sampling. The proposed approach "learns" a fast surrogate meta-model using a limited number of ABM evaluations and approximates the nonlinear relationship between ABM inputs (initial conditions and parameters) and outputs. Performance is evaluated on the (Brock and Hommes, 1998) asset pricing model and the "Islands" endogenous growth model (Fagiolo and Dosi, 2003). Results demonstrate that machine learning surrogates obtained using the proposed iterative learning procedure provide a quite accurate proxy of the true model and dramatically reduce the computation time necessary for large scale parameter space exploration and calibration.

Keywords: agent based model, calibration, machine learning, surrogate, meta-model

JEL Classification: C15, C52, C63

Suggested Citation

Lamperti, Francesco and Roventini, Andrea and Sani, Amir, Agent-Based Model Calibration Using Machine Learning Surrogates (October 3, 2017). LEM Working Paper. Available at SSRN: https://ssrn.com/abstract=2943297 or http://dx.doi.org/10.2139/ssrn.2943297

Francesco Lamperti (Contact Author)

Scuola Superiore Sant'Anna di Pisa - Institute of Economics and LEM ( email )

Institute of Economics
Piazza Martiri della Liberta, n. 33
Pisa, Pisa 56127
Italy

Fondazione Eni Enrico Mattei (FEEM) ( email )

C.so Magenta 63
Milano, 20123
Italy

Andrea Roventini

Scuola Superiore Sant'Anna di Pisa - Laboratory of Economics and Management (LEM) ( email )

Piazza Martiri della Liberta', 33-I-56127
Pisa
Italy

Observatoire Fran├žais des Conjonctures Economiques (OFCE)

69 Quai d'Orsay
Paris 75004
France

Amir Sani

Simudyne Ltd ( email )

35 Luke Street
London, London EC2A4LH
United Kingdom

Scuola Superiore Sant'Anna di Pisa - Laboratory of Economics and Management (LEM)

Biblioteca Scuola Superiore Sant'Anna
Piazza Martiri della Liberta, n. 33
Pisa, 56127
Italy

Register to save articles to
your library

Register

Paper statistics

Downloads
187
Abstract Views
805
rank
162,643
PlumX Metrics