Assistant Professor, ML / AI / IoT
University of Georgia
Fei Dou is an Assistant Professor in the School of Computing at the University of Georgia. She obtained her Ph.D. degree from the Department of Computer Science and Engineering at the University of Connecticut in Laboratory of Machine Learning & Health Informatics, working on Machine Learning (ML) / Artificial Intelligence (AI) in the Internet of Things (IoT) / Cyber-Physical Systems (CPS) supervised by Prof. Jinbo Bi. Previously, she was working on Underwater Acoustic Sensor Networks (UWASN) in Tianjin University supervised by Prof. Zhigang Jin.
Fei's research centers on developing new methodologies in AI/ML to enhance the Efficiency, Privacy, and Scalability of the IoT/CPS. Specifically, Fei is mainly working on Location-based Services (LBS), Edge Computing, and Remote Sensing Imagery Analysis by developing new methods from the perspectives of Reinforcement Learning, Federated Learning, and Contrastive Learning. Fei's investigations encompass various IoT domains, including location-based services, natural disaster damage detection, intelligent agriculture, recommender systems, and wireless sensor networks. Fei’s work has been published in highly-selective and high-impact conferences and journals such as NeurIPS, IJCAI, UbiComp/IMWUT, IEEE IOT-J, and she has served as a reviewer at top-notch conferences and journals including NeurIPS, ICLR, AAAI, IJCAI, ECAI, IEEE IOT-J, Neurocomputing, Sensors, etc.
Fei's research lies in analyzing and resolving challenges associated with Machine Intelligence of Ubiquitous Computing in Distributed Systems, including:
Recruiting: Looking for highly-motivated Ph.D. students to join my group! If you are interested in working on machine learning/deep learning/reinforcement learning/federated learning/contrastive learning, please send me an email with your CV and transcript @ fei dot dou AT uga dot edu.
[2024/09] One paper is accepted in IEEE Transactions on Mobile Computing 2024.
[2024/08] One paper is accepted in MobiCom Picasso Workshop 2024.
[2024/08] One paper is accepted in UbiComp/IMWUT 2024.
[2024/04] Received IAI Research Award.
[2024/01] Two papers are accepted in IEEE ICC 2024, one of the two IEEE Communications Society's flagship conferences.
[2023/10] I am thrilled and honored to be recognized as the Collegian Winner for the 2023 Women of Innovation Academic Innovation and Leadership in Connecticut! Please check out the latest news: patch, 8newsnow, UConnToday.
[2023/09] Our papers (two) are accepted by NeurIPS 2023! Congratulations to all!
[2023/09] Please check our new vision paper on AGI in the IOT.
[2023/08] I’m officially joining the School of Computing at the University of Georgia as a Tenure-Track Assistant Professor.
[2023/06] I am honored to be selected as the finalist of 2023 Women of Innovation(WOI) from Connecticut Technology Council!
[2023/05] Our paper is accepted by IEEE IOT-J 2023!
[2023/05] Our papers (two) are submitted to NeurIPS 2023.
[2023/04] Our paper is accepted by IJCAI 2023!
[2023/01] Our papers are submitted to IJCAI 2023.
[2022/11] I am invited to give talks on "Special Topics on Reinforcement Learning and its Applications" at the University of Michigan-Dearborn.
[2022/10] Our paper is submitted to IEEE Internet of Things Journal (IOT-J).
[2022/08] Our paper is submitted to IEEE Internet of Things Journal (IOT-J).
* 15 projects/papers, as the first author (project leader) or the second author (main contributor)
Latent Orthonormal Contrastive Learning in Disaster Damage Assessment Using Paired Remote Sensing Imagery
Fei Dou, Cameron Cianci, Jinbo Bi
Submitted
A latent orthonormal contrastive learning approach is proposed to handle the high-resolution
satellite/aerial imagery.
Bi-Subspace Saliency Detection
Jin Lu, Fei Dou, Chun-Hsi Huang
CCWC 2017 [paper]
A novel bi-subspace data-driven saliency detection model is proposed by consider the problem from subspace analysis
to characterize the background and foreground.
On-Device Indoor Positioning: A Federated Reinforcement Learning Approach with Heterogeneous Devices
Fei Dou, Jin Lu, Tan Zhu, Jinbo Bi
IOT-J 2023 [paper]
A personalized federated learning (FL) for reinforcement learning (RL) is proposed to automatically learn
environmental dynamics by client-environment interactions via RL and cope with the diversity of client devices
and their non-identical data distributions via personalized FL.
A Bisection Reinforcement Learning Approach to 3-D Indoor Localization
Fei Dou, Jin Lu, Tingyang Xu, Chun-Hsi Huang, and Jinbo Bi
IOT-J 2021 [paper]
A bisection reinforcement learning method is proposed to bisect the search space in a hierarchy from the entire
building down to a prespecified distance scale to the object position, and a unified framework for single-floor,
multifloor, and 3-D indoor localization is proposed.
Top-Down Indoor Localization with Wi-Fi Fingerprints using Deep Q-Network
Fei Dou, Jin Lu, Zigeng Wang, Xia Xiao, Jinbo Bi, Chun-Hsi Huang
MASS 2018 [paper]
A top-down searching method using a deep Q-network agent is proposed to tackle environment dynamics in
indoor positioning with Wi-Fi fingerprints, by formulating the indoor localization
problem as a Markov Decision Process rather than a typical classification or regression problem.
Mobilizing Personalized Federated Learning via Random Walk Stochastic ADMM
Ziba Parsons, Fei Dou, Houyi Du, Jin Lu
To Appear in NeurIPS 2023, [paper]
A Smart Narrow Down Approach based on Machine Learning for Indoor localization
Sahibzada Umair, Fei Dou, Tughrul Arslan
Under Review by IEEE Internet of Things Journal (IOT-J)
A Narrow down approach has been presented for indoor localization that involves in coarse and
accurate positioning phases, where training points selection, area division and overlapping strategies have been
presented to reduce the uncertainty.
Customized Positional Encoding to Combine Static and Time-varying Data in Robust Representation Learning for Crop Yield Prediction
Qinqing Liu, Fei Dou, Meijian Yang, Ezana Amdework, Guiling Wang, Jinbo Bi
IJCAI 2023,
[paper]
Polyhedron Attention Module: Learning Adaptive-order Interactions
Tan Zhu, Fei Dou, Xinyu Wang, Jinbo Bi
To Appear in NeurIPS 2023
Identifying Interactions among Categorical Predictors with Monte-Carlo Tree Search
Tan Zhu, Fei Dou, Chloe Becquey, Jinbo Bi
In Preparation, [paper]
A Deep Reinforcement Learning-based Path Planning for Wireless Rechargeable Sensor Network with Mobile Charger
Mainak Mondal, Fei Dou, Jinbo Bi, Song Han
In Preparation
On-demand Pipelined MAC for Multi-hop Underwater Wireless Sensor Networks
Fei Dou, Zheng Peng
WuWNet 2015,
[paper]
An on-demand pipeline is established to enable data transmission from the source to the destination over multiple
hops in a short time, where a special acknowledgement mechanism is designed to guarantee the reliability of
the communication with low overheads.
The Multi-channel MAC Protocol for High Performance Underwater Sensor Networks
Fei Dou, Zhigang Jin, Yao Zhang, Yishan Su
Journal of Harbin Engineering University 2015, CWSN 2013, [paper]
A multi-channel MAC protocol is proposed to tackle the spatial-temporal uncertainty by constructing the reservation model of the control channel using Markov Chain.
Motion Prediction Based MAC for Underwater Wireless Sensor Networks
Yishan Su, Zhigang Jin, Zixin Liu, Fei Dou*
Journal of Electronics & Information Technology 2013, [paper]
The motion model of underwater nodes is established to address the multiple access problem in underwater mobile networks,
with the aid of an AutoRegression (AR) mobility prediction algorithm.
WSF-MAC: A Weight-based Spatially Fair MAC Protocol for Underwater Sensor Networks
Fei Dou, Zhigang Jin, Yishan Su, Jian Liu
CECNet 2012, [paper]
A weight-based spatially fair MAC protocol is proposed to address the problem of spatial unfairness in UWSNs.
In recognition of her ability to motivate and effectively teach students, Fei was among the only few students who received the "Provost Recognition of Teaching Excellence" award from UCONN in 2017.
Giving lectures on basics of Markov Decision Process and Reinforcement Learning, including Bellman Equation, Value Iteration, Policy Iteration, Sampling Policy, Exploitation and Exploration, Model-based Learning, Model-free Approach, Monte Carlo Method, Temporal Difference, off-policy Q-Learning, on-policy SARSA, etc. Giving talks on Applications of Reinforcement Learning in real-world scenarios.
Download my CV here.
Fei Dou's research interests lie in analyzing and resolving challenges associated with Machine Intelligence of Ubiquitous Computing in Distributed Systems, including: