Shiyi graduated from the University of Chinese Academy of Sciences, majoring in Computer Science for her master’s and Statistics during her undergraduate.

Her master’s thesis was ‘The Extraction of Latent Spatiotemporal Patterns from Neural Signals,’ where she developed a deep learning framework to process noisy high-dimensional signals and revealed the latent spatiotemporal dynamics in neuron populations. Additionally, she was a quantitative consultant and data science researcher at a financial company, where she modelled stock time series using various factors. Her previous research experience also included inverse problems in bio-images, motion correction in low-resolution videos, optimization in complex network delivery, wave analysis in widefield imaging, deep learning in visual questioning answer of multimodal data (bachelor’s thesis), etc. She also enjoys attending mathematics, finance, and computer competitions in her spare time.

Shiyi is interested in how to develop data-driven methods to reduce manpower investment and build automatic, effective, and robust machine learning models efficiently for various real-world applications. She is excited to delve into next-generation statistical models for monitoring and improving water quality.

Apart from academic life, she enjoys going to the gym, reading, traveling by car with friends, visiting museums, and attending concerts (of almost all kinds – she can’t live without music).