Speaker：Dr. Zhigang Yao (Department of Statistics and Data Science, National University of Singapore)

  • Event Date: 2025-12-11
  • Speaker:  /  Host:


Topic：Manifold Fitting and a New Pipeline: CellScope

Speaker：Dr. Zhigang Yao (Department of Statistics and Data Science, National University of Singapore)

Time：Dec 11 (Thursday) , 2025, 13:20-14:10 

Place： 4F-427, Assembly Building I

 

Abstract

In this talk, I will discuss recent progress on manifold fitting and its role in bridging geometry and data science. While many existing methods rely on restrictive assumptions, the manifold hypothesis—that data often lie near non-Euclidean structures—remains fundamental in modern statistical learning. I will highlight both theoretical insights and algorithmic challenges, drawing on recent works as well as ongoing research. In particular, I will briefly introduce CellScope, a high-performance cell atlas workflow with a tree-structured representation. Compared to established pipelines such as Seurat and Scanpy, CellScope substantially improves clustering accuracy, visualization clarity, computational efficiency, and algorithmic interpretability, while reducing dependence on hyperparameters across a wide range of single-cell datasets.

 

1/0