R&D Intern - Fairness in Deep Learning

About the role

This research internship focuses on developing methods to ensure neural network predictions remain independent of continuous sensitive variables such as age or income. The intern will work within Artefact's research center to measure and mitigate algorithmic bias, moving beyond traditional binary or categorical fairness approaches. The role involves implementing existing regularization methods and designing new, differentiable bias-reduction techniques.

Responsibilities

• Implement existing bias-reduction methods based on dependency measures
• Establish a robust protocol to compare fairness and performance trade-offs
• Design and experiment with new bias-reduction methods for continuous variables
• Develop reproducible code using open-source libraries like PyTorch and scikit-learn

Required skills

• Deep Learning
• Machine Learning
• Statistics
• Applied Mathematics
• PyTorch
• scikit-learn

Qualifications

• Master's degree student in Applied Mathematics, Statistics, or Machine Learning
• Enrolled in a university or engineering school

About the Company

Artefact is an international data services company specializing in data transformation consulting. With 2500 employees, the firm bridges the gap between data and business to deliver tangible results through AI technologies and agile methods.