Understanding and Improving Fairness-Accuracy Trade-offs in Multi-Task Learning

As multi-task models gain popularity in a wider range of machine learning applications, it is becoming increasingly important for practitioners to understand the fairness implications associated with those models. Most existing fairness literature focuses on learning a single task more fairly, while how ML fairness interacts with multiple tasks in the joint learning setting is largely under-explored. In this paper, we are concerned with how group fairness (e.g., equal opportunity, equalized odds) as an ML fairness concept plays out in the multi-task scenario. In multi-task learning, several tasks are learned jointly to exploit task correlations for a more efficient inductive transfer. This presents a multi-dimensional Pareto frontier on (1) the trade-off between group fairness and accuracy with respect to each task, as well as (2) the trade-offs across multiple tasks. We aim to provide a deeper understanding on how group fairness interacts with accuracy in multi-task learning, and we show that traditional approaches that mainly focus on optimizing the Pareto frontier of multi-task accuracy might not perform well on fairness goals. We propose a new set of metrics to better capture the multi-dimensional Pareto frontier of fairness-accuracy trade-offs uniquely presented in a multi-task learning setting. We further propose a Multi-Task-Aware Fairness (MTA-F) approach to improve fairness in multi-task learning. Experiments on several real-world datasets demonstrate the effectiveness of our proposed approach.

[1]  Aaron Roth,et al.  Convergent Algorithms for (Relaxed) Minimax Fairness , 2020, ArXiv.

[2]  L. Khan,et al.  A Primal-Dual Subgradient Approach for Fair Meta Learning , 2020, 2020 IEEE International Conference on Data Mining (ICDM).

[3]  Indre Zliobaite,et al.  On the relation between accuracy and fairness in binary classification , 2015, ArXiv.

[4]  Sen Wu,et al.  Understanding and Improving Information Transfer in Multi-Task Learning , 2020, ICLR.

[5]  Maya R. Gupta,et al.  Training Well-Generalizing Classifiers for Fairness Metrics and Other Data-Dependent Constraints , 2018, ICML.

[6]  Krishna P. Gummadi,et al.  Fairness Constraints: Mechanisms for Fair Classification , 2015, AISTATS.

[7]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Elliot Meyerson,et al.  Pseudo-task Augmentation: From Deep Multitask Learning to Intratask Sharing - and Back , 2018, ICML.

[9]  Ed H. Chi,et al.  Practical Compositional Fairness: Understanding Fairness in Multi-Component Recommender Systems , 2019 .

[10]  Han Zhao,et al.  Conditional Learning of Fair Representations , 2019, ICLR.

[11]  Rishabh K. Iyer,et al.  A Reweighted Meta Learning Framework for Robust Few Shot Learning , 2020, ArXiv.

[12]  Dragomir Anguelov,et al.  Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout , 2020, NeurIPS.

[13]  Carlos Eduardo Scheidegger,et al.  Certifying and Removing Disparate Impact , 2014, KDD.

[14]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[15]  Kristian Lum,et al.  An algorithm for removing sensitive information: Application to race-independent recidivism prediction , 2017, The Annals of Applied Statistics.

[16]  S. Levine,et al.  Gradient Surgery for Multi-Task Learning , 2020, NeurIPS.

[17]  Jasbir S. Arora,et al.  Survey of multi-objective optimization methods for engineering , 2004 .

[18]  Bernhard Schölkopf,et al.  A Kernel Method for the Two-Sample-Problem , 2006, NIPS.

[19]  Luca Oneto,et al.  Taking Advantage of Multitask Learning for Fair Classification , 2018, AIES.

[20]  Krishna P. Gummadi,et al.  Fairness Constraints: A Flexible Approach for Fair Classification , 2019, J. Mach. Learn. Res..

[21]  Jonathan Baxter,et al.  A Model of Inductive Bias Learning , 2000, J. Artif. Intell. Res..

[22]  Jon M. Kleinberg,et al.  On Fairness and Calibration , 2017, NIPS.

[23]  Toon Calders,et al.  Building Classifiers with Independency Constraints , 2009, 2009 IEEE International Conference on Data Mining Workshops.

[24]  Ross B. Girshick,et al.  Fast R-CNN , 2015, 1504.08083.

[25]  John Langford,et al.  A Reductions Approach to Fair Classification , 2018, ICML.

[26]  Guillermo Sapiro,et al.  Minimax Pareto Fairness: A Multi Objective Perspective , 2020, ICML.

[27]  Toniann Pitassi,et al.  Learning Fair Representations , 2013, ICML.

[28]  Ed H. Chi,et al.  Transfer of Machine Learning Fairness across Domains , 2019, ArXiv.

[29]  Emily Denton,et al.  Detecting Bias with Generative Counterfactual Face Attribute Augmentation , 2019, ArXiv.

[30]  Linda F. Wightman LSAC National Longitudinal Bar Passage Study. LSAC Research Report Series. , 1998 .

[31]  Yu Zhang,et al.  A Survey on Multi-Task Learning , 2017, IEEE Transactions on Knowledge and Data Engineering.

[32]  Toniann Pitassi,et al.  Learning Adversarially Fair and Transferable Representations , 2018, ICML.

[33]  Zhe Zhao,et al.  Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts , 2018, KDD.

[34]  Jitendra Malik,et al.  Which Tasks Should Be Learned Together in Multi-task Learning? , 2019, ICML.

[35]  Chen Zhao,et al.  Rank-Based Multi-task Learning for Fair Regression , 2019, 2019 IEEE International Conference on Data Mining (ICDM).

[36]  Andrew McCallum,et al.  Ask the GRU: Multi-task Learning for Deep Text Recommendations , 2016, RecSys.

[37]  Roberto Cipolla,et al.  Multi-task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[38]  James Y. Zou,et al.  Multiaccuracy: Black-Box Post-Processing for Fairness in Classification , 2018, AIES.

[39]  Zhe Zhao,et al.  Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations , 2017, ArXiv.

[40]  Adam Tauman Kalai,et al.  Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings , 2016, NIPS.

[41]  Toon Calders,et al.  Classifying without discriminating , 2009, 2009 2nd International Conference on Computer, Control and Communication.

[42]  Allison Woodruff,et al.  Putting Fairness Principles into Practice: Challenges, Metrics, and Improvements , 2019, AIES.

[43]  Alexander D'Amour,et al.  Underspecification Presents Challenges for Credibility in Modern Machine Learning , 2020, J. Mach. Learn. Res..

[44]  Han Zhao,et al.  Inherent Tradeoffs in Learning Fair Representation , 2019, NeurIPS.

[45]  Xiaogang Wang,et al.  Deep Learning Face Attributes in the Wild , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[46]  Ed H. Chi,et al.  Toward a better trade-off between performance and fairness with kernel-based distribution matching , 2019, ArXiv.

[47]  Maya R. Gupta,et al.  Satisfying Real-world Goals with Dataset Constraints , 2016, NIPS.

[48]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[49]  Jason Weston,et al.  A unified architecture for natural language processing: deep neural networks with multitask learning , 2008, ICML '08.

[50]  Ananth Balashankar,et al.  What is Fair? Exploring Pareto-Efficiency for Fairness Constrained Classifiers , 2019, ArXiv.

[51]  Nathan Srebro,et al.  Equality of Opportunity in Supervised Learning , 2016, NIPS.

[52]  Guillermo Sapiro,et al.  Fairness With Minimal Harm: A Pareto-Optimal Approach For Healthcare , 2019, ArXiv.

[53]  Aditya Krishna Menon,et al.  The cost of fairness in binary classification , 2018, FAT.

[54]  Qiang Yang,et al.  An Overview of Multi-task Learning , 2018 .

[55]  Sebastian Ruder,et al.  An Overview of Multi-Task Learning in Deep Neural Networks , 2017, ArXiv.

[56]  Ed H. Chi,et al.  Fairness in Recommendation Ranking through Pairwise Comparisons , 2019, KDD.