Simulating human intuitive decisions by Q-learning

Abstract

Simulations have become a great tool for research in the natural sciences. However their potential has not been reached far enough in the social sciences. This is in part due to the difficulty in simulating human decision making and reproducing human-like behavior. Recent advances in neo-classical decision making have defined specific differences between the decision making capabilities of rational agents and humans as well as speculations into the cause. Presented is a Q-learning model for simulating human-like decision making based upon the intuition deliberation model proposed by psychologists Kahneman and Tversky. The model is tested against the classic economic bargaining game. In this game humans and rational agents consistently converge onto distinctly different strategies. Our experiments show that a selfish agent defers from the strategy of the rational agent and is more similar to human strategy.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 0 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.