General mechanisms for making decisions?

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Psychiatry

Background:

• Reward-guided behavior is often explained by computational decision-making models.
• These models emphasize expectation representation and prediction error-driven revision of future expectations.
• Frontal cortex and striatum are consistently implicated in these processes across species.

Purpose of the Study:

• To review the computational framework of decision-making.
• To discuss the neural underpinnings, particularly frontal cortex and striatum involvement.
• To explore the potential for analogous processes in non-reward-guided learning.

Main Methods:

• Review of existing literature on computational decision-making.
• Analysis of neuroimaging and behavioral studies in humans, monkeys, and rats.
• Synthesis of findings regarding prediction error signaling and its neural basis.

Main Results:

• A computational framework effectively models reward-guided decision-making.
• Prediction errors, crucial for learning, are signaled by frontal cortex and striatum.
• Evidence suggests these computational mechanisms may apply to broader learning contexts.

Conclusions:

• Computational models provide a unified framework for understanding decision-making and learning.
• Frontal cortex and striatum play key roles in processing prediction errors.
• The principles of expectation and prediction error may underlie diverse cognitive functions.