Researcher
Machine Learning Group,
Microsoft Research Asia (MSRA)
Google Scholar |
Microsoft Research |
Github
Zhihu
Weibo |
Wechat public account
Contact: Building 2, No. 5 DanLing Street, Haidian District, Beijing, China.
jindong.wang (at) microsoft.com
News
-
Internship: I'm looking for highly-motivated students to collaborate or
for internship. (The new internship period starts in 2021/05.) Please send me your CVs if you're interested. Keep an eye on this article for more detail.
-
202010: Our latest work Learning Causal Semantic Representation for Out-of-Distribution Prediction is released on ArXiv.
-
202009: My PhD thesis has been awarded the Excellent doctoral thesis by Chinese Academy of Sciences (中国科学院优秀博士学位论文奖)! Check this link.
-
202007: Paper Learning to Match Distributions for Domain Adaptation has been released on ArXiv.
-
2020-04-20: Paper Joint Partial Optimal Transport for Open Set Domain Adaptation has been accepted by IJCAI-20.
-
2020-04: Our paper Deep Subdomain Adaptation Network for Image Classification has been accepted by IEEE Trans. on Neural Networks and Learning Systems (TNNLS).
-
2020-04: Our paper FedHealth: A Federated Transfer Learning Framework for Wearable Healthcare has been accepted by IEEE Intelligent Systems (FedHealth was the Best Application Paper in IJCAI-19 federated learning workshop).
-
2020-02: Our paper Reliable Weighted Optimal Transport for Unsupervised Domain Adaptation has been accepted by CVPR 2020.
-
2019-12: Give a detailed introduction to transfer learning for Tsinghua University's EE graduates: [Class photo].
-
2019-10: Paper Cross-dataset activity recognition via adaptive spatial-temporal transfer learning has been accepted by UbiComp 2020.
Short Bio
I'm currently a researcher at machine learning group, Microsoft Research Asia (MSRA). Before joining MSRA, I obtained by Ph.D. from Institute of Computing Technology, Chinese Academy of Sciences in June, 2019. My Ph.D. thesis is mainly on transfer learning algorithms and applications in ubiquitous computing. In 2018/04--2018/08m, I was a visitor of Prof. Qiang Yang's group at Hong Kong University of Science and Technology (HKUST). At MSRA, my research interest is mainly about generalization problems, including but not limited to: transfer learning, few-shot learning, meta learning, adaptive learning, and related applications.
[Research] [Education] [Publications] [Talks] [Projects] [Awards & Service] [Skill] [Person]
Research
My main interest is the generalization-related problems. Generalization is perhaps the most fundamental and mysterious aspects of machine learning. Briefly speaking, it is the game of data, models, and optimizers, where each one of them plays a key role in helping a learning machine achieve better generalization abilities. Therefore, that machine can be non-trivially applied to a new environment with new tasks and datasets.
Such triumph would require efforts from many research areas, including but not limited to: deep learning, transfer learning, domain adaptation/generalization, out-of-distribution detection, meta learning, few-shot learning, adversarial robustness, adaptative learning, and so on. At the same time, these algorithms should be working with several applications to deal with real-world challenges, such as activity recognition, computer vision, financial technology, etc.