Active and Passive Learning in Agent-based Financial Markets
Author(s) -
Blake LeBaron
Publication year - 2010
Publication title -
eastern economic journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.276
H-Index - 22
eISSN - 1939-4632
pISSN - 0094-5056
DOI - 10.1057/eej.2010.53
Subject(s) - rationality , financial market , yield (engineering) , active learning (machine learning) , computer science , microeconomics , economics , finance , artificial intelligence , materials science , political science , law , metallurgy
This short note compares and contrasts two forms of learning which are present in most agent-based financial markets. First, passive learning refers to a form of “as if rationality” where wealth accumulates on strategies which have done relatively well. Second active learning refers to the active switching of agents across strategies. Most heterogeneous agent markets contain some form of both these types of learning. From what we know so far the dynamics of each may be quite different, and may yield a rich and complex joint dynamic.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom