Risk preference as an outcome of evolutionarily adaptive learning mechanisms: An evolutionary simulation under...

Area of Science:

• Cognitive Science
• Behavioral Economics
• Computational Neuroscience

Background:

• Cognitive and learning mechanisms underpin complex behaviors.
• Reinforcement learning (RL) models use distinct rules for positive and negative reward prediction errors.
• Risk preference exhibits both domain-specific and domain-general characteristics.

Purpose of the Study:

• To investigate the evolution of learning bias in relation to risk preference.
• To explore how different learning rules contribute to domain-specific and domain-general risk behaviors.
• To model the adaptive mechanisms underlying complex decision-making under risk.

Main Methods:

• Simulated the evolution of positive and negative learning rates in agents.
• Exposed agents to diverse risky environments with varying reward structures.
• Analyzed evolved learning rates and resulting behavioral patterns.

Main Results:

• The positive learning rate consistently evolved to be higher than the negative learning rate.
• Evolved agents demonstrated flexible reward-seeking and risk-averse behaviors based on task type.
• Simulated agents exhibited behavioral patterns consistent with prospect theory.

Conclusions:

• Evolved learning biases can explain the dual nature of risk preference (domain-specific and domain-general).
• A higher positive than negative learning rate facilitates adaptive behavior in varied risky contexts.
• The proposed framework of innate learning constraints offers insights into complex behavioral phenomena.