Choice-theoretic foundations of the divisive normalization model

Area of Science:

• Neuroscience
• Computational Neuroscience
• Decision Science

Background:

• Neuroscience suggests the brain employs utility-like calculations for decision-making.
• Divisive normalization is a computation implicated in these choice mechanisms.
• The underlying reasons and specific behaviors associated with normalization in choice remain unclear.

Purpose of the Study:

• To elucidate the rationale behind the brain's use of divisive normalization in choice.
• To define the specific behaviors consistent with normalization models.
• To establish a unified theory of choice behavior.

Main Methods:

• Proving a three-way equivalence theorem.
• Connecting the normalization model with an information-processing model.
• Linking these to an axiomatic characterization of behavior.

Main Results:

• Demonstrated equivalence between normalization, information-processing, and axiomatic models.
• Provided an optimality rationale for normalization in choice, balancing expected value and entropic cost.
• Characterized testable behaviors arising from normalization.

Conclusions:

• Unified three distinct models of choice into a single theoretical framework.
• Addressed the "how", "why", and "what" of brain choice mechanisms.
• Offered a comprehensive understanding of normalization's role in decision-making.