DocumentCode :
3720756
Title :
Modelling stock-market investors as Reinforcement Learning agents
Author :
Alvin Pastore;Umberto Esposito;Eleni Vasilaki
Author_Institution :
Department of Computer Science, University of Sheffield, United Kingdom
fYear :
2015
Firstpage :
1
Lastpage :
6
Abstract :
Decision making in uncertain and risky environments is a prominent area of research. Standard economic theories fail to fully explain human behaviour, while a potentially promising alternative may lie in the direction of Reinforcement Learning (RL) theory. We analyse data for 46 players extracted from a financial market online game and test whether Reinforcement Learning (Q-Learning) could capture these players behaviour using a riskiness measure based on financial modeling. Moreover we test an earlier hypothesis that players are “naíve” (short-sighted). Our results indicate that Reinforcement Learning is a component of the decision-making process. We also find that there is a significant improvement of fitting for some of the players when using a full RL model against a reduced version (myopic), where only immediate reward is valued by the players, indicating that not all players are naíve.
Keywords :
"Decision making","Investment","Maximum likelihood estimation","Learning (artificial intelligence)","Computational modeling","Economics","Psychology"
Publisher :
ieee
Conference_Titel :
Evolving and Adaptive Intelligent Systems (EAIS), 2015 IEEE International Conference on
Type :
conf
DOI :
10.1109/EAIS.2015.7368789
Filename :
7368789
Link To Document :
بازگشت