Yu Chen

Univ of Liverpool

Isterre, Université Grenoble Alpes

I’m currently a PhD candidate at Institute for Risk and Unceratinty, University of Liverpool; and also an ESR funded by the EU-Horizon 2020 & Marie Skłodowska-Curie Actions project URBASIS.

During my PhD, I am dedicated to developing robust Deep Learning-based computational frameworks against data problems (imprecise, limited, scarce, imbalanced or OOD data), and propagating associated uncertainty through computational probabilistic models in an efficient way. My contribution revolves around two aspects: (I) equiping DL models with uncertainty awareness, allowing for robustness; (II) incorporating DL with prior (physical) domain knowledge, allowing for fusion of knowledge.

Research interests

- **{Bayesian, knowledge-informed, Evidential, generative}** Deep Learning
- Stochastic modeling of uncertainties in engineering
- robustness of ML against data problems
- Imprecise probability
- Risk-based optimal decision making and cost-benefit analysis

news

2024	March Our submission to “PSAM17 & ASRAM2024” at Sendai, Japan has been accepted. Feb Our submission to “REC 2024” at Qinghua university Beijing has been accepted. Jan Our submission to “IVACS 2024” at Cambridge has been accepted.
2023	Nov I gave a talk at the “ESRA symoisum - Towards autonomous digital twin” at University of Strathclyde, Glasgow. . Sep Joined “Centre for Intelligent Infrastructure” at University of Strathclyde, Glasgow as an Research Assisstant for doing UQ in ML . July Our paper on Bayesian Augmented-Learning framework is now online . “The Vlog of URBASIS Spring school” is finally out June Presenting in 5th International Conference on Uncertainty Quantification in Computational Science and Engineering at Athens, Greece. May Attend URBASIS Spring School 2023 at Porquerolles, France April Take part in the 9th International School on Deep Learning at Bari, Italy.
2022	Sep Presenting in The 8th International Symposium on Reliability Engineering and Risk Management (ISRERM) at Hanover, Germany. Aug Presenting in 5th International Conference on Uncertainty Quantification in Computational Science and Engineering at Dublin, Ireland. July Attend the 6th International Gran Canaria School on Deep Learning (DeepLearn Summer) at Gran Canaria, Spain. Feb Start the secondment at ISTerre, UGA, Grenoble, France.
2021	June Take part in the ProbAI: Nordic Probabilistic AI School.

selected publications

A Bayesian Augmented-Learning Framework for Spectral Uncertainty Quantification of Incomplete Records of Stochastic Processes

Yu Chen, Edoardo Patelli, Benjamin Edwards, and Michael Beer

Mechanical Systems and Signal Processing, 2023

Bib HTML

@article{chen4405534bayesian,
  title = {A Bayesian Augmented-Learning Framework for Spectral Uncertainty Quantification of Incomplete Records of Stochastic Processes},
  author = {Chen, Yu and Patelli, Edoardo and Edwards, Benjamin and Beer, Michael},
  journal = {Mechanical Systems and Signal Processing},
  doi = {https://doi.org/10.1016/j.ymssp.2023.110573},
  url = {https://authors.elsevier.com/sd/article/S0888-3270(23)00481-8},
  year = {2023},
}