R&D Intern - Fairness in Deep Learning - Paris

The role

The intern will study neural networks to ensure prediction independence from continuous sensitive variables like age or income. The work involves implementing existing bias reduction methods and designing new methods using open source libraries.

What you will do

• Explore bias reduction methods for neural networks regarding continuous sensitive variables
• Implement existing methods such as regularizations based on dependence measures
• Set up a robust protocol to compare methods considering the performance and fairness tradeoff
• Design and experiment with new bias reduction methods adapted to continuous variables
• Use open source libraries like PyTorch and scikit-learn

Required qualifications

• BAC+4/BAC+5 in applied mathematics, statistics or machine learning from a university or engineering school
• Gap year internship or end-of-studies internship

About the company

Artefact is an international data services company specializing in data transformation consulting. Its 2500 employees use advanced AI technologies and agile methods to deliver business results.