Differentially Private and Fair Deep Learning: A Lagrangian Dual Approach.
Cuong TranFerdinando FiorettoPascal Van HentenryckPublished in: AAAI (2021)
Keyphrases
- deep learning
- differentially private
- lagrangian dual
- lower and upper bounds
- cutting plane algorithm
- linear programming
- differential privacy
- lagrangian relaxation
- unsupervised learning
- branch and bound
- linear program
- duality gap
- machine learning
- dual variables
- weakly supervised
- optimal solution
- integer programming
- lower bound
- privacy preserving
- upper bound
- dynamic programming
- object recognition
- active contours
- data sets
- high dimensional
- image processing