|
Weizhi Gao
I am a graduate student at the University of Chinese Academy
of Sciences, majoring in machine learning. Now I am advised by
Yingjie Tian.
I recieved my B.S. at the Beijing Normal University,
where I was majoring in mathmetics. In summer 2018, I went to
the Tufts University for a summer project with
full scholarship, studying basic coding.For now, I am looking for a Ph.D. If you have anything
to communicate, please contact me.
In my spare time, I love playing badminton. Though I am not very good at it, we can play together
as long as you want. Besides, I like singing as well. Please remind me if I do not notice I bother
others! :)
Email /
CV /
Github /
LinkedIn
|
|
|
Research Interests
I am mainly interested in reliable AI, leading machine learning to reliable tools that can be widely
deployed in the real scenario. To achieve this, I focus on uncertainty, robustness, and adaptability
of
deep learning. My long-term research goal is to develop AI to take responsibility for its decisions,
and
to enable AI to be used in accuracy-sensitive tasks (medical diagnosis, automatic driving, etc.).
- Model Uncertainty: Calibration
- Model Robustness: Out of Distribution Detection, Adversarial Attack
- Model Adaptability: Transfer Learning, Few Shot Learning
|
Research Experience
The Analysis of Luminosity Encoding Rule
- State Key Laboratory of Cognitive Neural Science and Learning, Beijng Normal University,
May, 2018 - May, 2019
- Topic: The Analysis of Luminosity Encoding Rule
- Contribution: Analyze the luminosity encoding rule in V1 cortex with neural network tools
|
Word Segmentation and Named Entity Recognition for Chinese Electronic Medical Record
- School of Mathematics Sciences, Beijing Normal University, Aug, 2019 - Jun, 2020
- Topic: Extract unstructured information from Chinese electronic medical records
- Contribution: Survey and implement traditional methods and deep learning methods on
Chinese
electronic medical records
|
A Survey of Transfer Learning
- School of Mathematics Sciences, UCAS, Sep, 2021 - Nov, 2021
- Topic: A Survey of Transfer Learning
- Contribution:Survey for existing popular transfer learning methods and the application of
transfer
learning in different tasks.
|
|
Two-stage Training Strategy Combined with Neural Network for Segmentation of
Internal
Mammary Artery Graft
- School of Mathematics Sciences, UCAS, Dec, 2021 - Mar, 2022
- Contribution: Construct a novel dataset, and propose an effective two-stage training
strategy for
internal mammary artery graft segmentation
- Publication: Biomedical Signal Processing and Control
|
|
Improving Long-Tailed Classification by Disentangled Variance Transfer
- School of Mathematics Sciences, UCAS, Mar, 2022 - Sep, 2022
- Contribution: Propose a disentangled variance transfer method, DisVar, to improve the
effectiveness
of knowledge transfer in long-tailed learning
- Publication: Under review in Journal of Internet of Things
|
Serives
- Serve as a reviewer for International Conference on Learning Representations 2023 (ICLR)
- Serve as a reviewer for ACM International Conference on Web Search and Data Mining 2023 (WSDM)
|
Awards
UCAS
- Academic Scholarship of School of Mathematics Science, UCAS, 2022
- Academic Scholarship of School of Mathematics Science, UCAS, 2021
- Academic Scholarship of School of Mathematics Science, UCAS, 2020
BNU
- Second Prize of Scholarship of Beijing Normal University, 2019
- Full scholarship for Tufts University summer school, 2018
- Second Prize of Beijing Normal University Mathematical Modeling Contest, 2018
- Second Prize of Scholarship of Beijing Normal University, 2018
- Second Prize of Scholarship of Beijing Normal University, 2017
- First Prize of New Student Scholarship (top 5%), 2016
- Second Prize of Scholarship of Beijing Normal University, 2016
|
|
|
|