Learning Fair Representations via Rate-Distortion Maximization
Chowdhury, Somnath Basu Roy, Chaturvedi, Snigdha
–arXiv.org Artificial Intelligence
Text representations learned by machine learning models often encode undesirable demographic information of the user. Predictive models based on these representations can rely on such information, resulting in biased decisions. We present a novel debiasing technique, Fairness-aware Rate Maximization (FaRM), that removes protected information by making representations of instances belonging to the same protected attribute class uncorrelated, using the rate-distortion function. FaRM is able to debias representations with or without a target task at hand. FaRM can also be adapted to remove information about multiple protected attributes simultaneously. Empirical evaluations show that FaRM achieves state-of-the-art performance on several datasets, and learned representations leak significantly less protected attribute information against an attack by a non-linear probing network.
arXiv.org Artificial Intelligence
Aug-25-2022
- Country:
- Oceania > Australia
- North America
- Dominican Republic (0.04)
- United States
- Texas > Travis County
- Austin (0.04)
- Oregon > Multnomah County
- Portland (0.04)
- Minnesota > Hennepin County
- Minneapolis (0.14)
- Louisiana > Orleans Parish
- New Orleans (0.04)
- Georgia > Fulton County
- Atlanta (0.04)
- Colorado > Boulder County
- Boulder (0.04)
- California > Los Angeles County
- Long Beach (0.04)
- Texas > Travis County
- Canada > British Columbia
- Vancouver (0.04)
- Europe
- Slovenia (0.04)
- Czechia > Prague (0.04)
- United Kingdom > Scotland
- City of Edinburgh > Edinburgh (0.04)
- Spain > Catalonia
- Barcelona Province > Barcelona (0.04)
- Middle East > Republic of Türkiye
- Istanbul Province > Istanbul (0.04)
- Iceland > Capital Region
- Reykjavik (0.04)
- Denmark > Capital Region
- Copenhagen (0.04)
- Belgium > Brussels-Capital Region
- Brussels (0.04)
- Asia
- Genre:
- Research Report (0.82)
- Technology:
- Information Technology
- Data Science > Data Mining (1.00)
- Communications (1.00)
- Artificial Intelligence
- Representation & Reasoning (1.00)
- Natural Language > Text Processing (1.00)
- Machine Learning
- Statistical Learning (0.67)
- Performance Analysis > Accuracy (0.46)
- Information Technology