Machine learning for economic ABMs: From fast calibration to AI-agents
This event will be in hybrid format but will not be recorded.
Artificial Intelligence (AI) is transforming science, driving a paradigm shift through the use of machine learning (ML) algorithms that automatically improve with more data and/or more computational power. Agent-Based Models (ABMs), with their computational nature, are uniquely positioned to lead this transformation also within economics. The Applied Research Team at the Bank of Italy’s IT Department has contributed to this shift through three complementary research lines. First, efficient, well-engineered, and user-friendly ABMs are of foundational importance. Aldo Glielmo will present two state-of-the-art open-source packages: ABCredIT and BeforeIT, designed to provide solid frameworks for ABM usage and experimentation. Second, ML techniques make calibrating ABM parameters on large datasets increasingly feasible. Aldo will discuss Black-IT, a dedicated calibration engine, and explore how Reinforcement Learning (RL) can enhance calibration efficiency. Finally, integrating AI-software agents into ABMs may relieve modellers from manually defining complex behavioural rules. Aldo will illustrate the introduction of RL agents in traditional ABMs and discuss the promises and challenges of incorporating agents based on Large Language Models (LLMs).

About the speaker

Aldo Glielmo is a research scientist in the Bank of Italy’s IT research unit, working at the intersection of Artificial Intelligence and Economics. He holds a PhD in Physics from King’s College London. Prior to joining the Bank of Italy, he conducted research at the Alan Turing Institute and the International School for Advanced Studies.
Date: 26 February 2025, 14:30
Venue: Manor Road Building, Manor Road OX1 3UQ
Venue Details: Seminar Room G and online via Zoom
Speaker: Dr Aldo Glielmo (Bank of Italy)
Organising department: Institute for New Economic Thinking
Organisers: Dorothy Nicholas (INET Oxford), Francois Lafond (INET Oxford)
Organiser contact email address: complexity@inet.ox.ac.uk
Host: Prof. Doyne Farmer
Part of: INET Complexity Economics Seminars
Booking required?: Required
Booking url: https://www.inet.ox.ac.uk/events/machine-learning-for-economic-abms
Audience: Public
Editors: Dorothy Nicholas, Fiona Burbage