Qi Qi @ The University of Iowa
Education
I am a graduate student from machine learning and optimization group of University of Iowa under the supervision of Prof. Tianbao Yang. Before joined UIowa, I received master degree from University of Science and Technology of China under the supervision of Prof. Thomas Weise and bachelor degree from Qingdao University of Technology .
Research Interests
Rooted in machine learning and distributionally robust learning, my research focuses on developing efficient and robust algorithms for large-scale deep learning tasks, such as deep metric learning, contrastive learning, data imbalance problems with heavy-tailed label distributions and fairness under awareness/unawareness.
News
- [11/22] Try our new package ABSGD for handling the general imbanaced classification by just modifying the loss and optimizer to ABLoss, ABSGD/ABAdam ! Tutorial!
- [10/22] Our ABSGD achieve 1st in ResNet50 (rank 4th of 16 in total) on the iWildCam dataset in Standford Wilds out of distribution competition! Try our Code!
- [08/22] Second amazing intern journey at Netflix, work with Shervin Ardeshir and M. Hossein Taghavi
- [01/22] Give an invited talk in 2022 at UberAI about AUPRC maximization.
Packages and Competitions
- I am the author of CIDFLIX, ABSGD package and a contributor of LibAUC (impelemented the AUPRC algorithm).
- Top 1 in ResNet50 (4th/16), iWildCam multi-class species classification, by Standford, 10/2022, [Code]
Manuscripts
- [ New! ] Fairness via Adversarial Attribute Neighbourhood Robust Learning. [Paper][Code will be released soon]
Qi Qi, Shervin Ardershir, Yi Xu, Tianbao Yang.
Publications
- Improving Identity-Robustness for Face Models [Paper][Code]
Qi Qi, Shervin Ardershir
The short version to appear at (ICML 2023 SCIS Workshop)
- Stochastic Constrained DRO with a Complexity Independent of Sample Size. [Paper][Code]
Qi Qi, Jiameng Lyu, Kung-sik Chan, Er-Wei Bai, Tianbao Yang.
Transactions on Machine Learning Research (TMLR, 2023)
- ABSGD: Attentional Biased Stochastic Gradient for Imbalanced Classification. [Paper] [Code]
Qi Qi, Yi Xu, Rong Jin, Wotao Yin, Tianbao Yang.
Transactions on Machine Learning Research (TMLR, 2023)
- AUPRC: Stochastic Optimization of Area Under Precision-Recall Curve for Deep Learning with Provable Convergence [Paper][Code Package] [Poster] [Slides]
Qi Qi, Youzhi Luo, Zhao Xu, Shuiwang Ji, Tianbao Yang.
Proceedings of the 35th Conference on Neural Information Processing Systems (NeurIPS 2021)
- RECOVER: A Practical Online Method for Distributionally Deep Robust Optimization. [Paper][Code][Poster] [Slides]
Qi Qi, Zhishuai Guo, Yi Xu, Rong Jin, Tianbao Yang.
Proceedings of the 35th Conference on Neural Information Processing Systems (NeurIPS 2021)
- Variance-Reduced Off-Policy Memory-Efficient Policy Search [Paper][Code]
Daoming Lyu, Qi Qi, Mohammad Ghavamzadeh, Hengshuai Yao, Tianbao Yang, Bo Liu.
Offline Reinforcement Learning Workshop at 34th-Neural Information Processing Systems (Neurips 2020 Workshop)
- A Simple and Effective Framework for Pairwise Deep Metric Learning. [Paper][Code][Poster] [Video]
Qi Qi, Yan Yan, Xiaoyu Wang, Tianbao Yang.
16th European Conference on Computer Vision (ECCV 2020)
- Stochastic optimization for DC functions and non-smooth non-convex regularizers with non-asymptotic convergence. [Paper][Code]
Yi Xu, Qi Qi, Qihang Lin, Rong Jin, Tianbao Yang.
Proceedings of the 36th International Conference on Machine Learning (ICML 2019).
- Modeling optimization algorithm runtime behavior and its applications. [Paper]
Qi Qi, Thomas Weise, and Bin Li.
Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO 2017).
- Optimization algorithm behavior modeling: A study on the traveling salesman problem[C] [Paper][Slides]
Qi Qi, Thomas Weise, and Bin Li.
Tenth International Conference on Advanced Computational Intelligence ( ICACI 2018).
- Automatically discovering clusters of algorithm and problem instance behaviors as well as their causes from experimental data, algorithm setups, and instance features[J] [Paper]
Thomas Weise, Xiaofeng Wang, Qi Qi, Ke Tang, Bin Li.
Applied Soft Computing, 2018, 73: 366-382.
Experiences
- Netflix, Computer Vision Intern, June - August, 2021/2022
• Improved the robustness of face models by proposing an conditional inverse density (CID) robust weighting method.
• Implemented an internal python package CIDFLIX , including the proposed CID method and robust evaluations.
• Submitted to CVPR2023 [Illustrations]
Teaching Assistant
Selected Awards
- First Prize of Graduate Scholarship, University of Science and Technology of China, Sep. 2017.
- Outstadning Graduate Awards, Qingdao Technology University, Jul. 2015.
- National Undergraduate Scholarship, Ministry of Education of China, Nov. 2012.
Services
2020 Fall Seminar
2021 Spring Seminar