Agent based modelling

From ACT Wiki
Jump to: navigation, search


A type of forecast modelling in which individuals or collective entities (such as organisations or groups) and their interaction with each other and their environment, are explicitly represented.

ABM assumes that participants in the system are autonomous agents, rather than responding in entirely predictable or passive ways, as assumed by some 'classical' forecasting models.

ABM attempts to integrate human and institutional behaviour - including 'irrational' behaviour - with a view to assessing their effects on the system as a whole.

Both mathematical and experimental approaches are required for ABM's development and application.

Bank of England's economic forecasting

"Andy Haldane, chief economist at the Bank of England... puts a lot of store in 'agent based modelling' (ABM), a computational approach that attempts to integrate the often capricious nature of human and institutional behaviour into forecasting models.
Let's hope it works; if it doesn't, populist denigration of expert opinion will only intensify further.
The alternative of government by gut instinct is not an appealing prospect."
The Treasurer magazine, February 2017, p13 - Jeremy Warner, assistant editor of The Daily Telegraph.

See also