November 28, 2025Experimental EconomicsOpen Access

Strategies in the multi-armed bandit

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Experiment shows reinforcement learning strategies fit behaviors in multi-armed bandit tasks, indicating diversity in responses.

Most subjects favor reinforcement learning in multi-armed bandit tasks, showing variability in individual strategies.
The analysis involved a between-subjects experiment with different configurations of bandit problems.
Data suggests that biased strategies adapt effectively to incorporate behavioral tendencies of subjects.
Findings emphasize the need for tailored approaches in understanding decision-making in multi-armed bandit scenarios.

Hudja et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark