When overconfident agents slow down collective learning - Université Paris 1 Panthéon-Sorbonne
Journal Articles SIMULATION: Transactions of The Society for Modeling and Simulation International Year : 2012

When overconfident agents slow down collective learning

Abstract

This paper presents a model of influence where agents' beliefs are based on an objective reality, such as the properties of an environment. The perception of the objective reality is not direct: all agents know is that the more correct a belief, the more successful the actions that are deduced from this belief. (A pair of agents can influence each other when )Agents can influence eachother by pair when they perform a joint action. They are not only defined by individual beliefs, but also idyosynchratic confidence in their belief - this means that they are not all willing to (engage in action with) act with agents with a different belief and to be influenced by them. We show here that the distribution of confidence in the group has a huge impact on the speed and quality of collective learning and in particular that a small number of overconfident agents can prevent the whole group frow learning properly.
Fichier principal
Vignette du fichier
VersFinSimulationJuillet11-2.pdf (549.92 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-00623966 , version 1 (15-09-2011)

Identifiers

Cite

Juliette Rouchier, Emily Tanimura. When overconfident agents slow down collective learning. SIMULATION: Transactions of The Society for Modeling and Simulation International, 2012, 88 (1), pp.33-49. ⟨10.1177/0037549711428948⟩. ⟨hal-00623966⟩
328 View
582 Download

Altmetric

Share

More