Mitigating Unwanted Biases with Adversarial Learning

ABSTRACT

Machine learning is a tool for building models that accurately represent input training data. When undesired biases concerning demographic groups are in the training data, well-trained models will reflect those biases. We present a framework for mitigating such biases by including a variable for the group of interest and simultaneously learning a predictor and an adversary. The input to the network X, here text or census data, produces a prediction Y, such as an analogy completion or income bracket, while the adversary tries to model a protected variable Z, here gender or zip code. The objective is to maximize the predictor's ability to predict Y while minimizing the adversary's ability to predict Z. Applied to analogy completion, this method results in accurate predictions that exhibit less evidence of stereotyping Z. When applied to a classification task using the UCI Adult (Census) Dataset, it results in a predictive model that does not lose much accuracy while achieving very close to equality of odds (Hardt, et al., 2016). The method is flexible and applicable to multiple definitions of fairness as well as a wide range of gradient-based learning models, including both regression and classification tasks.

References

1. Asuncion, A., and Newman, D. 2007. Uci machine learning repository.
2. Beutel, A.; Chen, J.; Zhao, Z.; and Chi, E. H. 2017. Data decisions and theoretical implications when adversarially learning fair representations. arXiv preprint arXiv:1707.00075.
3. Bolukbasi, T.; Chang, K.-W.; Zou, J. Y.; Saligrama, V.; and Kalai, A. T. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. In Advances in Neural Information Processing Systems, 4349--4357.
4. Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 2014. Generative adversarial nets. In Advances in neural information processing systems, 2672--2680.
5. Hardt, M.; Price, E.; Srebro, N.; et al. 2016. Equality of opportunity in supervised learning. In Advances in Neural Information Processing Systems, 3315--3323.
6. Kingma, D., and Ba, J. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
7. Kleinberg, J.; Mullainathan, S.; and Raghavan, M. 2016. Inherent trade-offs in the fair determination of risk scores. arXiv preprint arXiv:1609.05807.
8. Lum, K., and Johndrow, J. 2016. A statistical framework for fair predictive algorithms. arXiv preprint arXiv:1610.08077.
9. Mikolov, T.; Sutskever, I.; Chen, K.; Corrado, G. S.; and Dean, J. 2013. Distributed representations of words and phrases and their compositionality. In Advances in neural information processing systems, 3111--3119.

Index Terms

• Computing methodologies

  • Artificial intelligence

    • Natural language processing

  • Machine learning

    • Learning paradigms

      • Multi-task learning

      • Supervised learning

        • Supervised learning by classification

    • Machine learning approaches

      • Neural networks

• Mathematics of computing

  • Mathematical analysis

    • Mathematical optimization

      • Continuous optimization

        • Convex optimization

  • Probability and statistics

    • Probabilistic inference problems

      • Maximum likelihood estimation

• Social and professional topics

  • User characteristics

    • Gender

    • Race and ethnicity

• Theory of computation

  • Design and analysis of algorithms

    • Mathematical optimization

      • Continuous optimization

        • Nonconvex optimization

  • Theory and algorithms for application domains

    • Machine learning theory

      • Reinforcement learning

        • Adversarial learning