We seek a talented full-time computational associate in research to work at Duke University on an exciting new project at the interface of machine learning and biology. The T.Dunn and Tadross labs are working together to transform how we tackle Parkinson’s disease, via development and application of cutting edge computational and molecular tools. The associate will advance new deep neural networks for computer vision and use them to analyze new high-dimensional datasets of real biological systems.

Specific Work Performed

-Use custom lab deep learning software (dannce -- www.github.com/spoonsso/dannce) to analyze existing video datasets. This will include running dannce over videos via slurm on the Duke high performance compute cluster.

-Optimize dannce performance by running hyperparameter and network architecture searches.

- While working with the codebase, improve the dannce pipeline via refactoring and integration of pipeline subcomponents. Contribute to dannce documentation and public releases on GitHub.

-Use existing lab scripts to analyze dannce pose tracking data, with the end goal to identify animal action type clusters and assess how these clusters change across experimental conditions. The position requires fluency in any programming language and familiarity with any deep learning API. No prior job experience is required, but applicants must have or expect (by May 2021) a bachelor’s degree. The project is a good fit for a computational scientist looking to work with biological data or for a biologist with a strong computational background, and as such there are no formal requirements for the type of degree.

The position will provide opportunities to co-author publications and build a productive portfolio for applying to graduate programs or industry jobs in any field valuing extensive experience with machine learning and big data.

Duke University is an Affirmative Action/Equal Opportunity Employer committed to providing employment opportunity without regard to an individual's age, color, disability, gender, gender expression, gender identity, genetic information, national origin, race, religion, sex, sexual orientation, or veteran status.

Duke aspires to create a community built on collaboration, innovation, creativity, and belonging. Our collective success depends on the robust exchange of ideas-an exchange that is best when the rich diversity of our perspectives, backgrounds, and experiences flourishes. To achieve this exchange, it is essential that all members of the community feel secure and welcome, that the contributions of all individuals are respected, and that all voices are heard. All members of our community have a responsibility to uphold these values.