Instrumental Variable-Driven Domain Generalization with Unobserved Confounders

Domain generalization (DG) aims to learn from multiple source domains a model that can generalize well on unseen target domains. Existing DG methods mainly learn the representations with invariant marginal distribution of the input features, however, the invariance of the conditional distribution of the labels given the input features is more essential for unknown domain prediction. Meanwhile, the existing of unobserved confounders which affect the input features and labels simultaneously cause spurious correlation and hinder the learning of the invariant relationship contained in the conditional distribution. Interestingly, with a causal view on the data generating process, we find that the input features of one domain are valid instrumental variables for other domains. Inspired by this finding, we propose an instrumental variable-driven DG method (IV-DG) by removing the bias of the unobserved confounders with two-stage learning. In the first stage, it learns the conditional distribution of the input features of one domain given input features of another domain. In the second stage, it estimates the relationship by predicting labels with the learned conditional distribution. Theoretical analyses and simulation experiments show that it accurately captures the invariant relationship. Extensive experiments on real-world datasets demonstrate that IV-DG method yields state-of-the-art results.

[1]  Kun Kuang,et al.  Learning Decomposed Representations for Treatment Effect Estimation , 2023, IEEE Transactions on Knowledge and Data Engineering.

[2]  B. Ooi,et al.  IDEAL: Toward High-efficiency Device-Cloud Collaborative and Dynamic Recommendation System , 2023, ArXiv.

[3]  Kun Kuang,et al.  DUET: A Tuning-Free Device-Cloud Collaborative Parameters Generation Framework for Efficient Device Model Generalization , 2022, WWW.

[4]  Lanfen Lin,et al.  Collaborative Semantic Aggregation and Calibration for Federated Domain Generalization , 2021, IEEE Transactions on Knowledge and Data Engineering.

[5]  Lin Liu,et al.  Learning Causal Representations for Robust Domain Adaptation , 2020, IEEE Transactions on Knowledge and Data Engineering.

[6]  Lanfen Lin,et al.  Knowledge Distillation-based Domain-invariant Representation Learning for Domain Generalization , 2023, IEEE Transactions on Multimedia.

[7]  Kun Kuang,et al.  Domain Generalization via Contrastive Causal Learning , 2022, ArXiv.

[8]  Hanrui Wu,et al.  Multiple Graphs and Low-Rank Embedding for Multi-Source Heterogeneous Domain Adaptation , 2022, ACM Trans. Knowl. Discov. Data.

[9]  Kun Kuang,et al.  Personalizing Intervened Network for Long-tailed Sequential User Behavior Modeling , 2022, ArXiv.

[10]  Lanfen Lin,et al.  Label-Efficient Domain Generalization via Collaborative Exploration and Generalization , 2022, ACM Multimedia.

[11]  Wang Lu,et al.  Domain-invariant Feature Exploration for Domain Generalization , 2022, Trans. Mach. Learn. Res..

[12]  Hao Wang,et al.  Bayesian Invariant Risk Minimization , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  E. Bareinboim,et al.  Causal Transportability for Visual Recognition , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Baochun Li,et al.  OrphicX: A Causality-Inspired Latent Variable Model for Interpreting Graph Neural Networks , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Lanfen Lin,et al.  Attention-based Cross-Layer Domain Alignment for Unsupervised Domain Adaptation , 2022, Neurocomputing.

[16]  Lanfen Lin,et al.  Domain-Specific Bias Filtering for Single Labeled Domain Generalization , 2021, International Journal of Computer Vision.

[17]  Lap-Pui Chau,et al.  Variational Disentanglement for Domain Generalization , 2021, Trans. Mach. Learn. Res..

[18]  Lanfen Lin,et al.  Auto IV: Counterfactual Prediction via Automatic Instrumental Variable Decomposition , 2021, ACM Trans. Knowl. Discov. Data.

[19]  Heng Tao Shen,et al.  Adversarial Entropy Optimization for Unsupervised Domain Adaptation , 2021, IEEE Transactions on Neural Networks and Learning Systems.

[20]  Jonas Peters,et al.  A Causal Framework for Distribution Generalization , 2020, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[21]  Jindong Wang Introduction to domain generalization , 2022 .

[22]  Peng Cui,et al.  Deep Stable Learning for Out-Of-Distribution Generalization , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[23]  Y. Qiao,et al.  Domain Generalization with MixStyle , 2021, ICLR.

[24]  Hao Wang,et al.  Generative Interventions for Causal Learning , 2020, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[25]  Amit Sharma,et al.  Domain Generalization using Causal Matching , 2020, ICML.

[26]  N. Meinshausen,et al.  Anchor regression: Heterogeneous data meet causality , 2018, Journal of the Royal Statistical Society: Series B (Statistical Methodology).

[27]  Shiqi Wang,et al.  Domain Generalization for Medical Imaging Classification with Linear-Dependency Regularization , 2020, NeurIPS.

[28]  David Zhang,et al.  Guide Subspace Learning for Unsupervised Domain Adaptation , 2020, IEEE Transactions on Neural Networks and Learning Systems.

[29]  Judy Hoffman,et al.  Learning to Balance Specificity and Invariance for In and Out of Domain Generalization , 2020, ECCV.

[30]  Lequan Yu,et al.  Learning from Extrinsic and Intrinsic Supervisions for Domain Generalization , 2020, ECCV.

[31]  Timothy M. Hospedales,et al.  Learning to Generate Novel Domains for Domain Generalization , 2020, ECCV.

[32]  Eric P. Xing,et al.  Self-Challenging Improves Cross-Domain Generalization , 2020, ECCV.

[33]  Kun Zhang,et al.  A Causal View on Robustness of Neural Networks , 2020, NeurIPS.

[34]  Xi Peng,et al.  Learning to Learn Single Domain Generalization , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[35]  Sunita Sarawagi,et al.  Efficient Domain Generalization via Common-Specific Low-Rank Decomposition , 2020, ICML.

[36]  Tao Xiang,et al.  Deep Domain-Adversarial Image Generation for Domain Generalisation , 2020, AAAI.

[37]  Z. Geng,et al.  Causal Inference , 2020, Engineering.

[38]  Yuchen Guo,et al.  From Brain Science to Artificial Intelligence , 2020 .

[39]  Nan Duan,et al.  Progress in Neural NLP: Modeling, Learning, and Reasoning , 2020, Engineering.

[40]  Kui Ren,et al.  Adversarial Attacks and Defenses in Deep Learning , 2020, Engineering.

[41]  Jiashi Feng,et al.  Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation , 2020, ICML.

[42]  Kun Zhang,et al.  Domain Adaptation As a Problem of Inference on Graphical Models , 2020, NeurIPS.

[43]  Tatsuya Harada,et al.  Domain Generalization Using a Mixture of Multiple Latent Domains , 2019, AAAI.

[44]  Bohyung Han,et al.  Learning to Optimize Domain Specific Normalization for Domain Generalization , 2019, ECCV.

[45]  MengChu Zhou,et al.  Unsupervised Domain Adaptation With Adversarial Residual Transform Networks , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[46]  Tongliang Liu,et al.  Domain Generalization via Entropy Regularization , 2020, NeurIPS.

[47]  Daniel C. Castro,et al.  Domain Generalization via Model-Agnostic Learning of Semantic Features , 2019, NeurIPS.

[48]  David Lopez-Paz,et al.  Invariant Risk Minimization , 2019, ArXiv.

[49]  Arthur Gretton,et al.  Kernel Instrumental Variable Regression , 2019, NeurIPS.

[50]  Andrew Bennett,et al.  Deep Generalized Method of Moments for Instrumental Variable Analysis , 2019, NeurIPS.

[51]  Yuchen Zhang,et al.  Bridging Theory and Algorithm for Domain Adaptation , 2019, ICML.

[52]  Fabio Maria Carlucci,et al.  Domain Generalization by Solving Jigsaw Puzzles , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[53]  Yongxin Yang,et al.  Episodic Training for Domain Generalization , 2019, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[54]  Wei Zhou,et al.  Feature-Critic Networks for Heterogeneous Domain Generalization , 2019, ICML.

[55]  Junmo Kim,et al.  Learning Not to Learn: Training Deep Neural Networks With Biased Data , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[56]  Bo Wang,et al.  Moment Matching for Multi-Source Domain Adaptation , 2018, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[57]  Liang Lin,et al.  Larger Norm More Transferable: An Adaptive Feature Norm Approach for Unsupervised Domain Adaptation , 2018, 2019 IEEE/CVF International Conference on Computer Vision (ICCV).

[58]  Swami Sankaranarayanan,et al.  MetaReg: Towards Domain Generalization using Meta-Regularization , 2018, NeurIPS.

[59]  D. Tao,et al.  Deep Domain Generalization via Conditional Invariant Adversarial Networks , 2018, ECCV.

[60]  Alex ChiChung Kot,et al.  Domain Generalization with Adversarial Feature Learning , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[61]  Silvio Savarese,et al.  Generalizing to Unseen Domains via Adversarial Data Augmentation , 2018, NeurIPS.

[62]  Dacheng Tao,et al.  Domain Generalization via Conditional Invariant Representations , 2018, AAAI.

[63]  Siddhartha Chaudhuri,et al.  Generalizing Across Domains via Cross-Gradient Training , 2018, ICLR.

[64]  Tatsuya Harada,et al.  Maximum Classifier Discrepancy for Unsupervised Domain Adaptation , 2017, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[65]  Yongxin Yang,et al.  Learning to Generalize: Meta-Learning for Domain Generalization , 2017, AAAI.

[66]  Michael I. Jordan,et al.  Conditional Adversarial Domain Adaptation , 2017, NeurIPS.

[67]  Yongxin Yang,et al.  Deeper, Broader and Artier Domain Generalization , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[68]  Kevin Leyton-Brown,et al.  Deep IV: A Flexible Approach for Counterfactual Prediction , 2017, ICML.

[69]  Sethuraman Panchanathan,et al.  Deep Hashing Network for Unsupervised Domain Adaptation , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[70]  Michael I. Jordan,et al.  Deep Transfer Learning with Joint Adaptation Networks , 2016, ICML.

[71]  Bernhard Schölkopf,et al.  Domain Adaptation with Conditional Transferable Components , 2016, ICML.

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

[73]  Mengjie Zhang,et al.  Domain Generalization for Object Recognition with Multi-task Autoencoders , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[74]  Michael I. Jordan,et al.  Learning Transferable Features with Deep Adaptation Networks , 2015, ICML.

[75]  Bernhard Schölkopf,et al.  Multi-Source Domain Adaptation: A Causal View , 2015, AAAI.

[76]  Victor S. Lempitsky,et al.  Unsupervised Domain Adaptation by Backpropagation , 2014, ICML.

[77]  Bernhard Schölkopf,et al.  Domain Adaptation under Target and Conditional Shift , 2013, ICML.

[78]  Jieping Ye,et al.  Multisource domain adaptation and its application to early detection of fatigue , 2012, TKDD.

[79]  Bernhard Schölkopf,et al.  A Kernel Two-Sample Test , 2012, J. Mach. Learn. Res..

[80]  Gilles Blanchard,et al.  Generalizing from Several Related Classification Tasks to a New Unlabeled Sample , 2011, NIPS.

[81]  Koby Crammer,et al.  A theory of learning from different domains , 2010, Machine Learning.

[82]  Neil D. Lawrence,et al.  Dataset Shift in Machine Learning , 2009 .

[83]  Joshua D. Angrist,et al.  Mostly Harmless Econometrics: An Empiricist's Companion , 2008 .

[84]  W. Newey,et al.  Instrumental variable estimation of nonparametric models , 2003 .

[85]  Illtyd Trethowan Causality , 1938 .

[86]  Philip G. Wright,et al.  The tariff on animal and vegetable oils , 1928 .