Timezone: »

Pushing the Accuracy-Fairness Tradeoff Frontier with Introspective Self-play
Jeremiah Liu · Krishnamurthy Dvijotham · Jihyeon Lee · Quan Yuan · Martin Strobel · Balaji Lakshminarayanan · Deepak Ramachandran

Improving the accuracy-fairness frontier of deep neural network (DNN) models is an important problem. Uncertainty-based active learning (AL) can potentially improve the frontier by preferentially sampling underrepresented subgroups to create a more balanced training dataset. However, the quality of uncertainty estimates from modern DNNs tend to degrade in the presence of spurious correlations and dataset bias, compromising the effectiveness of AL for sampling tail groups. In this work, we propose $Introspective Self-play$ (ISP), a simple approach to improve the uncertainty estimation of a deep neural network under dataset bias, by adding an auxiliary $Introspection$ task requiring a model to predict the bias for each data point in addition to the label. We show that ISP provably improves the bias-awareness of the model representation and the resulting uncertainty estimates. On two real-world tabular and language tasks,ISP serves as a simple “plug-in” for AL model training, consistently improving both the tail-group sampling rate and the final accuracy-fairness trade-off frontier of popular AL methods.

#### Author Information

##### Jihyeon Lee (Google Research, Stanford University)

Jihyeon is an incoming software engineer at Google Research. She graduated with a BS and MS in Computer Science from Stanford University, where she was a researcher in the Sustain Lab advised by Professor Stefano Ermon and in the Vision & Learning Lab advised by Professor Fei-Fei Li. She is interested in semi-supervised learning and AI methods to solve problems at the intersection of environment, policy, and people.

##### Balaji Lakshminarayanan (Google Brain)

Balaji Lakshminarayanan is a research scientist at Google Brain. Prior to that, he was a research scientist at DeepMind. He received his PhD from the Gatsby Unit, University College London where he worked with Yee Whye Teh. His recent research has focused on probabilistic deep learning, specifically, uncertainty estimation, out-of-distribution robustness and deep generative models. Notable contributions relevant to the tutorial include developing state-of-the-art methods for calibration under dataset shift (such as deep ensembles and AugMix) and showing that deep generative models do not always know what they don't know. He has co-organized several workshops on "Uncertainty and Robustness in deep learning" and served as Area Chair for NeurIPS, ICML, ICLR and AISTATS.