Analysis of acoustic signaling and collective movement in animal groups using machine learning
1 PhD / 1 Postdoctoral Position
We are seeking a postdoc or PhD student to join an exciting new research project investigating how vocal signals mediate collective movement and decision-making in animal groups. The postdoc / student will develop and apply machine learning approaches to analyze acoustic and movement data recorded simultaneously from multiple members of animal groups in the wild.
Background. Recent studies of collective behavior in animal groups have shown that coordinated group movement can emerge if individuals obey surprisingly simple behavioral rules. However, many species have evolved sophisticated communication systems known to play a role in the coordination of movement. Because animals can actively vary when, how often, and what types of signals they produce, this introduces the possibility for individuals to dynamically shape their signaling behavior to alter their interactions with others. Understanding the role such signals play in shaping movement decisions, and conversely how movement dynamics affect signal production, is key to understanding the mechanisms and evolution of collective movement across a wide range of systems.
Simultaneous tracking data on the movements and vocalizations of multiple individuals within moving animal groups offers a powerful new perspective on the interplay between communication and collective movement. Processing, analyzing, and interpreting such behavioral ‘big data’ requires integrating biological insights with modern computational approaches, raising important challenges and opportunities.
Position. The successful applicant will join an interdisciplinary and international research team studying signaling and collective movement across multiple species of social mammals. Their main role will be to develop and apply machine learning approaches (e.g. convolutional neural networks) to detect, classify, and analyze vocalizations recorded on GPS / audio collars deployed on meerkats and spotted hyenas, using both supervised and unsupervised approaches. They will also collaborate with field biologists in analyzing combined movement / acoustic data to reveal the individual decision-making mechanisms that underlie coordinated group behaviors.
Qualifications. The ideal candidate will have both a quantitative / computational background and a strong motivation to tackle biological questions. To be considered for a postdoctoral position, candidates are required to have a PhD in any quantitative discipline including computer science, biology, physics, mathematics, etc. Highly motivated students with a Masters degree in a quantitative discipline are also welcome to apply to pursue this project as a PhD student. Applicants must have experience programming in Python, R, Matlab, or similar, and evidence of experience with software development will be positively viewed. Prior experience with machine learning (especially deep learning) is advantageous but not required, however an enthusiasm for diving into new computational approaches is essential. A collaborative spirit and the ability to work as part of an interdisciplinary team are also essential. The working language of the group is English, and German language skills are not a requirement. Women and members of underrepresented minorities are especially encouraged to apply.
Research community. The post doc or student will join the research group of Ariana Strandburg-Peshkin within the Department of Biology at the University of Konstanz. They will also be integrated within the newly-established Centre for the Advanced Study of Collective Behaviour, through which they will have the opportunity to interact with researchers in a range of fields including computer science, biology, psychology, and economics. Beyond Konstanz, they will work closely with an interdisciplinary and international team of collaborators involved in an emerging cross-species investigation of signaling and collective movement.
Advisor: The advisor for this project is Ariana Strandburg-Peshkin. For informal questions, please contact firstname.lastname@example.org