(CN) – Birds of a feather flock together, but research published Tuesday in the American Physics Institute journal Chaos looks for the leaders that determine where the seemingly headless group moves.
“Understanding the underlying coordination mechanism of these appealing phenomena helps us gain more cognition of the world where we live,” said author Duxin Chen, an assistant professor at China’s Southeast University in a statement.
In search of “an information theory-based method to reveal the causal relationship among individuals,” researchers from Southeast University in Nanjing and the China University of Mining and Technology banded together to analyze datasets collected from three pigeon flocks, with 10 birds each.
The data used in this study was originally collected by University of Konstanz bio-physicist Máté Nagy in a 2013 study aiming to understand dominance and leadership in swarm-like groups.
Researchers then fed their data on individual bird position, velocity, and acceleration in a k-nearest neighbor data mining algorithm, designed to recognize patterns in problem sets without specific parameters.
Computer science has been working toward a data-driven explanation of flock behavior for decades. Craig Reynold’s 1987 Boids program simulated “bird-oid objects” in bird-like flocks, based on the “famous three rules: avoid collision, velocity alignment, and cohesion to center.”
In this study, Chinese researchers challenge that while previous research offers “sufficiently intuitive explanations to the interaction mechanism and emergence phenomenon of coordinated behavior, the proposed simple interaction rules still lack the support of empirical evidence.”
Even as few as 10 individual birds create a complex system, with each individual at once responding to its neighbors and influencing their behavior.
“For pigeon flocks,” researchers noted, “the flight competition is intensive, which means the individuals are not likely to establish fixed causal connections with others. The interacting neighbors are easy to change when angular velocity is larger, since the constraints to keep cohesive exist.”
But one trend stuck out: the most influential birds fly closest to the center.
“Interestingly, the individuals closer to the mass center and the average velocity direction are more influential to others, which means location and flight direction are two factors that matter in their interactions,” Chen said.
In an email, Chen added that he hopes to see the research applied to better understand the collective motion of animal groups.
“The key factor is to calculate the mutual information in order to obtain causality,” Chen explained. “However, different kinds of collective motion have their own features, so it could be applied to human traffic patterns if the assumption of the algorithm is reasonable in the considered human mobility pattern.”
The researchers are also interested in seeing the algorithm tested against flocks of starlings and herds of large mammals, which behave differently in their own collectives.
Currently the team is using a similar model-free causality analysis to understand the collective behaviors of immune system cells.