Skip to main content
 

animal-soup: Automated animal behavioral classification using machine learning (2024)

Undergraduate: Caitlin Lewis


Faculty Advisor: Adam Hantman
Department: UNC Neuroscience Center


The Hantman Lab uses mouse models and a reach-to-grab behavioral task where a mouse has to grab a food pellet and eat it. Various perturbations can be performed, such as moving the food pellet or disrupting brain regions, to help us understand various aspects of skilled motor behavior. An imperative part of understanding the skilled movement during this task in mice is analyzing their behavior in parallel with the corresponding neural activity. However, hand-labeling components of the behavior manually can be very time intensive and often have varying results due to the subjective nature of human labeling. Thus, the use of machine learning techniques to automate this process will allow for better, more efficient analysis of skilled motor movement in the Hantman Lab that can have broader reaching impacts on the neuroscience community.