TY - GEN
T1 - Gradient Based Activations for Accurate Bias-Free Learning
AU - Kurmi, Vinod K.
AU - Sharma, Rishabh
AU - Sharma, Yash Vardhan
AU - Namboodiri, Vinay P.
PY - 2022/6/30
Y1 - 2022/6/30
N2 - Bias mitigation in machine learning models is imperative, yet challenging. While several approaches have been proposed, one view towards mitigating bias is through adversarial learning. A discriminator is used to identify the bias attributes such as gender, age or race in question. This discriminator is used adversarially to ensure that it cannot distinguish the bias attributes. The main drawback in such a model is that it directly introduces a trade-off with accuracy as the features that the discriminator deems to be sensitive for discrimination of bias could be correlated with classification. In this work we solve the problem. We show that a biased discriminator can actually be used to improve this bias-accuracy tradeoff. Specifically, this is achieved by using a feature masking approach using the discriminator's gradients. We ensure that the features favoured for the bias discrimination are de-emphasized and the unbiased features are enhanced during classification. We show that this simple approach works well to reduce bias as well as improve accuracy significantly. We evaluate the proposed model on standard benchmarks. We improve the accuracy of the adversarial methods while maintaining or even improving the unbiasness and also outperform several other recent methods.
AB - Bias mitigation in machine learning models is imperative, yet challenging. While several approaches have been proposed, one view towards mitigating bias is through adversarial learning. A discriminator is used to identify the bias attributes such as gender, age or race in question. This discriminator is used adversarially to ensure that it cannot distinguish the bias attributes. The main drawback in such a model is that it directly introduces a trade-off with accuracy as the features that the discriminator deems to be sensitive for discrimination of bias could be correlated with classification. In this work we solve the problem. We show that a biased discriminator can actually be used to improve this bias-accuracy tradeoff. Specifically, this is achieved by using a feature masking approach using the discriminator's gradients. We ensure that the features favoured for the bias discrimination are de-emphasized and the unbiased features are enhanced during classification. We show that this simple approach works well to reduce bias as well as improve accuracy significantly. We evaluate the proposed model on standard benchmarks. We improve the accuracy of the adversarial methods while maintaining or even improving the unbiasness and also outperform several other recent methods.
UR - http://www.scopus.com/inward/record.url?scp=85147653852&partnerID=8YFLogxK
U2 - 10.1609/aaai.v36i7.20687
DO - 10.1609/aaai.v36i7.20687
M3 - Chapter in a published conference proceeding
AN - SCOPUS:85147653852
T3 - Proceedings of the 36th AAAI Conference on Artificial Intelligence, AAAI 2022
SP - 7255
EP - 7262
BT - AAAI-22 Technical Tracks 7
PB - Association for the Advancement of Artificial Intelligence (AAAI)
T2 - 36th AAAI Conference on Artificial Intelligence, AAAI 2022
Y2 - 22 February 2022 through 1 March 2022
ER -