Student Project in Time Series Representation Learning


Time series analysis plays an important role in various industries such as medical informatics, healthcare, financial market, and climate modeling. Recent development in machine learning has promised to revolutionize predicting and classifying time sequences, by means of temporal convolutional networks and transformers. Although many state-of-the-art models have reached good performance in time series forecasting and classification, many time series tasks remain a challenge. Our research interest lies in time series representation learning, which aims to capture both contextual and temporal information at any resolution and generate representations to improve multiple downstream tasks (patient sequence similarity, etc.).

Inspired by recent contrastive representation learning in the field of computer vision, we hope to improve the performance of SOTA time series models by designing adaptive contrastive methods.


  • Strong motivation, and knowledge in deep learning/machine learning/statistics.
  • Master student in Mathematics/Computer Science/Data Science/Computational Science.
  • Experience with PyTorch and NLP/CV is an advantage.
  • Weekly meetings to discuss ideas and next steps.

How to apply

To apply, please send your CV, your MS/BS transcripts by email to

For CBB master students at ETH, please check this form .

For other students at ETH, it is also easier - you need be enrolled at UZH as a mobility student, see here . If you have any questions regarding ETH regulation, please contact Student Exchange Office: Dr. Francesca Broggi,, Tel +41 44 632 43 46.


[1] Gerald Woo, Chenghao Liu, Doyen Sahoo, Akshat Kumar, and Steven Hoi. “CoST: Contrastive learning of disentangled seasonal-trend representations for time series forecasting.” In International Conference on Learning Representations (ICLR), 2022.

[2] H. Yèche, G. Dresdner, F. Locatello, M. Hüser, and G. Rätsch, “Neighborhood Contrastive Learning Applied to Online Patient Monitoring.” In Proceedings of the 38th International Conference on Machine Learning (ICML), Jul. 2021, vol. 139, pp. 11964–11974.

[3] Cheng, Joseph Y., et al. “Subject-aware contrastive learning for biosignals.” arXiv preprint arXiv:2007.04871 (2020).

[4] Zang, Chengxi, and Fei Wang. “SCEHR: Supervised Contrastive Learning for Clinical Risk Prediction using Electronic Health Records.” arXiv preprint arXiv:2110.04943 (2021).

Xiaochen Zheng
PhD Student

Data science and machine learning.