Statistical analyses of Siamese fighting fish aggressive behavior
Project: analyze behavior of Siamese fighting fish (Betta splendens) as part of a collaboration between the Bendesky and Cunningham labs of the Zuckerman Institute (NeuroTheory Center)
Bettas are an ideal model for understanding how visual stimuli elicit aggressive behavior in a sex-specific manner (only males exhibit aggressive display). Raw data are side- and top-view videos (90 minutes/trial; 40 fps) of male-male and male-female pairs. We are using machine learning-based tracking software DeepLabCut (Mathis et al., 2018) to recognize 5 points on each fish’s body and return cartesian coordinates for each point. From this positional data, we’ve extracted several behavioral features (gill flaring, orientation, 2D trajectory). The student will:
- Develop and validate mathematical methods to extract additional behavioral data from DLC coordinates (e.g. 3D trajectory, fin extension);
- Define prevalent behavioral states based on these behaviors;
- Create and validate HMMs, GLM-HMMs, and other models for individual animals and pairs of animals (Calhoun et al., 2019) The novelty of this project is developing HMMs that incorporate the behavior of the partner animal as sensory feedback, taking into account both internal state and external cues in driving a behavioral outcome.
This is an UNPAID research project.
Faculty Advisor
- Professor: Andres Bendesky 1 2
- Department/School: Zuckerman Mind Brain Behavior Institute
- The diversity of animal behavior is fascinating, yet we know little about how behavior evolves. We take genetic, genomic, molecular, and neurobiological approaches to understand the mechanisms underlying the natural variation and evolution of behavior.
Project Timeline
- Earliest starting date: 3/1/2020
- End date:
- Number of hours per week of research expected during Spring 2020: ~10
- Number of hours per week of research expected during Summer 2020: ~20
Candidate requirements
- Skill sets:
- Knowledge of R and/or Python
- Familiarity with hidden markov models
- Interest in animal behavior
- Student eligibility: freshman, sophomore, junior, senior, master’s
- International students on F1 or J1 visa: eligible