{"type":"ScholarlyArticle","title":"A Unified Meta-Learning Framework for Dynamic Transfer Learning","authors":[{"name":"Jun Wu","slug":"jun-wu","given":"Jun","family":"Wu","affiliation":"University of Illinois Urbana-Champaign"},{"name":"Jingrui He","slug":"jingrui-he","given":"Jingrui","family":"He","affiliation":"University of Illinois Urbana-Champaign"}],"abstract":"Transfer learning refers to the transfer of knowledge or information from a relevant source task to a target task. However, most existing works assume both tasks are sampled from a stationary task distribution, thereby leading to the sub-optimal performance for dynamic tasks drawn from a non-stationary task distribution in real scenarios. To bridge this gap, in this paper, we study a more realistic and challenging transfer learning setting with dynamic tasks, i.e., source and target tasks are continuously evolving over time. We theoretically show that the expected error on the dynamic target task can be tightly bounded in terms of source knowledge and consecutive distribution discrepancy across tasks. This result motivates us to propose a generic meta-learning framework L2E for modeling the knowledge transferability on dynamic tasks. It is centered around a task-guided meta-learning problem with a group of meta-pairs of tasks, based on which we are able to learn the prior model initialization for fast adaptation on the newest target task. L2E enjoys the following properties: (1) effective knowledge transferability across dynamic tasks; (2) fast adaptation to the new target task; (3) mitigation of catastrophic forgetting on historical target tasks; and (4) flexibility in incorporating any existing static transfer learning algorithms. Extensive experiments on various image data sets demonstrate the effectiveness of the proposed L2E framework.","doi":"10.24963/ijcai.2022/496","url":"https://ijcai.org/proceedings/2022/496","pdf":"https://ijcai.org/proceedings/2022/0496.pdf","markdown":"https://ijcai.org/proceedings/2022/496.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/496","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"Machine Learning","track":"Main Track","number":496,"pages":{"first":3573,"last":3579},"keywords":["Machine Learning: Meta-Learning","Machine Learning: Multi-task and Transfer Learning"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38987012","length_sec":91,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38985155","length_sec":718,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Acuna et al., 2021] David Acuna, Guojun Zhang, Marc T. Law, and Sanja Fidler. f -domain adversarial learning: Theory and algorithms. In ICML, 2021.","title":"f -domain adversarial learning: Theory and algorithms","authors":"David Acuna, Guojun Zhang, Marc T Law, Sanja Fidler","year":"2021","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Al-Halah et al., 2017] Ziad Al-Halah, Rainer Stiefelhagen, and Kristen Grauman. Fashion forward: Forecasting vi- sual style in fashion. In ICCV, 2017.","title":"Fashion forward: Forecasting visual style in fashion","authors":"Ziad Al-Halah, Rainer Stiefelhagen, Kristen Grauman","year":"2017","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Ben-David et al., 2010] Shai Ben-David, John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira, and Jennifer Wortman Vaughan. A theory of learning from different domains. Machine learning, 2010.","title":"A theory of learning from different domains","authors":"Shai Ben-David, John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira, Jennifer Wortman Vaughan","year":"2010","venue":"Machine learning","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Bobu et al., 2018] Andreea Bobu, Eric Tzeng, Judy Hoff- man, and Trevor Darrell. Adapting to continuously shift- ing domains. In ICLR Workshop, 2018.","title":"Adapting to continuously shifting domains","authors":"Andreea Bobu, Eric Tzeng, Judy Hoffman, Trevor Darrell","year":"2018","venue":"ICLR Workshop","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Fallah et al., 2020] Alireza Fallah, Aryan Mokhtari, and Asuman Ozdaglar. On the convergence theory of gradient- based model-agnostic meta-learning algorithms. In AIS- TATS, 2020.","title":"On the convergence theory of gradientbased model-agnostic meta-learning algorithms","authors":"Alireza Fallah, Aryan Mokhtari, Asuman Ozdaglar","year":"2020","venue":"AIS-TATS","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Finn et al., 2017] Chelsea Finn, Pieter Abbeel, and Sergey Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In ICML, 2017.","title":"Model-agnostic meta-learning for fast adaptation of deep networks","authors":"Finn","year":"2017","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Finn et al., 2019] Chelsea Finn, Aravind Rajeswaran, Sham Kakade, and Sergey Levine. Online meta-learning. In ICML, 2019.","title":"Online meta-learning","authors":"Chelsea Finn, Aravind Rajeswaran, Sham Kakade, Sergey Levine","year":"2019","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Ganin et al., 2016] Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, Franc ¸ois Laviolette, Mario Marchand, and Victor Lempitsky. Domain-adversarial training of neural networks. JMLR, 17(1):2096-2030, 2016.","title":"Domain-adversarial training of neural networks","authors":"Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, Mario Franc ¸ois Laviolette, Victor Marchand, Lempitsky","year":"2016","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Ghifary et al., 2016] Muhammad Ghifary, David Balduzzi, W Bastiaan Kleijn, and Mengjie Zhang. Scatter compo- nent analysis: A unified framework for domain adaptation and domain generalization. TPAMI, 2016.","title":"Scatter component analysis: A unified framework for domain adaptation and domain generalization","authors":"Ghifary","year":"2016","venue":"TPAMI","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Gretton et al., 2012] Arthur Gretton, Karsten M Borgwardt, Malte J Rasch, Bernhard Schölkopf, and Alexander Smola. A kernel two-sample test. JMLR, 2012.","title":"A kernel two-sample test","authors":"Arthur Gretton, Karsten M Borgwardt, J Malte, Bernhard Rasch, Alexander Schölkopf, Smola","year":"2012","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[He et al., 2016] Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recog- nition. In CVPR, pages 770-778, 2016.","title":"Deep residual learning for image recognition","authors":"Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Hoffman et al., 2014] Judy Hoffman, Trevor Darrell, and Kate Saenko. Continuous manifold based adaptation for evolving visual domains. In CVPR, 2014.","title":"Continuous manifold based adaptation for evolving visual domains","authors":"Hoffman","year":"2014","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Kumar et al., 2020] Ananya Kumar, Tengyu Ma, and Percy Liang. Understanding self-training for gradual domain adaptation. In ICML, 2020.","title":"Understanding self-training for gradual domain adaptation","authors":"Ananya Kumar, Tengyu Ma, Percy Liang","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Li and Hoiem, 2017] Zhizhong Li and Derek Hoiem. Learning without forgetting. TPAMI, 2017.","title":"Learning without forgetting","authors":"Hoiem ; Li, Zhizhong Li, Derek Hoiem","year":"2017","venue":"TPAMI","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Liu et al., 2020] Hong Liu, Mingsheng Long, Jianmin Wang, and Yu Wang. Learning to adapt to evolving do- mains. NeurIPS, 2020.","title":"Learning to adapt to evolving domains","authors":"Liu","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Long et al., 2015] Mingsheng Long, Yue Cao, Jianmin Wang, and Michael Jordan. Learning transferable features with deep adaptation networks. In ICML, 2015.","title":"Learning transferable features with deep adaptation networks","authors":"Mingsheng Long, Yue Cao, Jianmin Wang, Michael Jordan","year":"2015","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Mansour et al., 2009] Yishay Mansour, Mehryar Mohri, and Afshin Rostamizadeh. Domain adaptation: Learning bounds and algorithms. In COLT, 2009.","title":"Domain adaptation: Learning bounds and algorithms","authors":"Yishay Mansour, Mehryar Mohri, Afshin Rostamizadeh","year":"2009","venue":"COLT","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Mohri and Medina, 2012] Mehryar Mohri and An- dres Munoz Medina. New analysis and algorithm for learning with drifting distributions. In ALT, 2012.","title":"New analysis and algorithm for learning with drifting distributions","authors":"Medina Mohri, Mehryar Mohri, Andres Munoz Medina","year":"2012","venue":null,"doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Pan and Yang, 2009] Sinno Jialin Pan and Qiang Yang. A survey on transfer learning. TKDE, 2009.","title":"A survey on transfer learning","authors":"Jialin Sinno, Qiang Pan, Yang","year":"2009","venue":"TKDE","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Peng et al., 2019] Xingchao Peng, Qinxun Bai, Xide Xia, Zijun Huang, Kate Saenko, and Bo Wang. Moment match- ing for multi-source domain adaptation. In ICCV, 2019.","title":"Moment matching for multi-source domain adaptation","authors":"Peng","year":"2019","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Rafailidis and Nanopoulos, 2015] Dimitrios Rafailidis and Alexandros Nanopoulos. Modeling users preference dy- namics and side information in recommender systems. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 46(6):782-792, 2015.","title":"Modeling users preference dynamics and side information in recommender systems","authors":"Dimitrios Rafailidis, Alexandros Nanopoulos","year":"2015","venue":"IEEE Transactions on Systems, Man, and Cybernetics: Systems","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Rosenstein et al., 2005] Michael T Rosenstein, Zvika Marx, Leslie Pack Kaelbling, and Thomas G Dietterich. To trans- fer or not to transfer. In NIPS workshop, 2005.","title":"To transfer or not to transfer","authors":"Zvika Michael T Rosenstein, Leslie Marx, Thomas G Pack Kaelbling, Dietterich","year":"2005","venue":"NIPS workshop","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Sun et al., 2019] Qianru Sun, Yaoyao Liu, Tat-Seng Chua, and Bernt Schiele. Meta-transfer learning for few-shot learning. In CVPR, 2019.","title":"Meta-transfer learning for few-shot learning","authors":"Sun","year":"2019","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Wang et al., 2020] Hao Wang, Hao He, and Dina Katabi. Continuously indexed domain adaptation. In ICML, 2020.","title":"Continuously indexed domain adaptation","authors":"Hao Wang, Hao He, Dina Katabi","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Wen et al., 2020] Junfeng Wen, Russell Greiner, and Dale Schuurmans. Domain aggregation networks for multi- source domain adaptation. In ICML, 2020.","title":"Domain aggregation networks for multisource domain adaptation","authors":"Junfeng Wen, Russell Greiner, Dale Schuurmans","year":"2020","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Wu and He, 2020] Jun Wu and Jingrui He. Continuous transfer learning with label-informed distribution align- ment. arXiv preprint arXiv:2006.03230, 2020.","title":"Continuous transfer learning with label-informed distribution alignment","authors":"He Wu, Jun Wu, Jingrui He","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Wu and He, 2021] Jun Wu and Jingrui He. Indirect invisible poisoning attacks on domain adaptation. In KDD, 2021.","title":"Indirect invisible poisoning attacks on domain adaptation","authors":"He Wu, Jun Wu, Jingrui He","year":"2021","venue":"KDD","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Zhang et al., 2019] Yuchen Zhang, Tianle Liu, Mingsheng Long, and Michael Jordan. Bridging theory and algorithm for domain adaptation. In ICML, 2019.","title":"Bridging theory and algorithm for domain adaptation","authors":"Yuchen Zhang, Tianle Liu, Mingsheng Long, Michael Jordan","year":"2019","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Zhao et al., 2018] Han Zhao, Shanghang Zhang, Guanhang Wu, José MF Moura, Joao P Costeira, and Geoffrey J Gordon. Adversarial multiple source domain adaptation. NeurIPS, 2018.","title":"Adversarial multiple source domain adaptation","authors":"Han Zhao, Shanghang Zhang, Guanhang Wu, M F José, Joao P Moura, Geoffrey J Costeira, Gordon","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":30,"raw":"[Zhou et al., 2017] Yao Zhou, Lei Ying, and Jingrui He. MultiC 2 : an optimization framework for learning from task and worker dual heterogeneity. In SDM, 2017.","title":"MultiC 2 : an optimization framework for learning from task and worker dual heterogeneity","authors":"Yao Zhou, Lei Ying, Jingrui He","year":"2017","venue":"SDM","doi":null,"url":null,"internal_url":null},{"position":31,"raw":"[Zhou et al., 2019] Yao Zhou, Lei Ying, and Jingrui He. Multi-task crowdsourcing via an optimization framework. TKDD, 13(3):1-26, 2019.","title":"Multi-task crowdsourcing via an optimization framework","authors":"Yao Zhou, Lei Ying, Jingrui He","year":"2019","venue":"TKDD","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}