Machine Learning Capable of Detecting Social Grooming Behavior in Wild Baboons through Acceleration Signals: Study Findings

Scientists from Swansea University and the University of Cape Town have tracked the social grooming behavior of wild baboons using accelerometers attached to collars.

The study published in the journal Royal Society for Open Scienceis the first to successfully calculate initialization budgets using this method, which opens up full scope for future research directions.

Using collars containing accelerometers built at Swansea University, the team recorded the activities of baboons in Cape Town, South Africa, identifying and measuring general activities such as resting, walking, foraging, running and also giving and receiving grooming.

A supervised machine learning algorithm was trained on accelerator data matching baboon video recordings and successfully recognized giving and receiving with high accuracy.

The team then applied their machine learning model to grooming data collected from 12 monkeys in order to continuously identify grooming and other behaviors throughout the day and night.

Lead author Dr Charlotte Christensen from the University of Zurich said: “We weren’t sure whether a sensor on a collar would be able to detect behavior involving such subtle movements, but it worked. Our findings have important implications for the study of social behavior in animals, Especially in non-human primates.”

Social grooming is one of the most important social behaviors in primates, and since the 1950s it has become a central focus of research in primate zoology.

Previously, scientists relied on direct observations to determine how many primates cared for each other, and while direct observations provide systematic data, it is sparse and not continuous, with the added limitation that researchers can only watch a few animals at a time. .

Technology such as the one used in this study has revolutionized the field of animal behavior research and allows for exciting new areas of research.

Senior author Dr Innes Fortbauer from Swansea University said: “This is something our team has wanted to do for years. Being able to collect and analyze ongoing grooming data in wild populations will allow researchers to re-examine long-standing questions and tackle new ones. Those related to the formation and maintenance of social bonds, as well On the mechanisms underlying social relationship, health, and physical fitness.