On the dimensionality of behavior.

Area of Science:

• Behavioral science
• Biophysics
• Computational neuroscience

Background:

• The field of "physics of behavior" is expanding, focusing on quantitative characterization of animal movements in naturalistic settings.
• High-dimensional data from animal movement studies present challenges in analysis.
• Identifying underlying low-dimensional structures is a key response to this data complexity.

Purpose of the Study:

• To provide a clearer definition of "dimensionality of behavior."
• To explore how dimensionality applies to both continuous animal trajectories and discrete behavioral states.
• To identify conditions under which behavioral dimensionality can be considered effectively infinite.

Main Methods:

• Conceptual analysis and definition of behavioral dimensionality.
• Examination of different representations of observable behavior (continuous trajectories, discrete states).
• Theoretical exploration of factors contributing to high or infinite dimensionality.

Main Results:

• Behavioral dimensionality can be understood through the lens of data structure, distinguishing between continuous and discrete state spaces.
• The framework allows for the identification of specific scenarios where behavioral complexity leads to effectively infinite dimensionality.
• Clarifies the quantitative approaches needed for analyzing complex animal behaviors.

Conclusions:

• A precise definition of behavioral dimensionality is crucial for interpreting high-dimensional animal movement data.
• Understanding dimensionality aids in developing more effective analytical tools for behavioral science.
• The concept of infinite dimensionality highlights the potential for emergent complexity in animal behavior.