Return to Colloquia & Seminar listing
Learning nonlinear structures from high-dimensional and noisy datasets
Mathematics of Data & Decisions| Speaker: | Xiucai Ding, UC Davis |
| Location: | 2112 MSB |
| Start time: | Tue, Oct 31 2023, 1:10PM |
In this talk, I will present recent findings pertaining to the comprehension of widely used unsupervised learning methods in the context of that a manifold is embedded in a low-dimensional Euclidean subspace and corrupted by high-dimensional noise. Specifically, I will delve into Diffusion Map (DM), Alternating Diffusion (AD) and Multiview Diffusion Map (MDM). This talk is based on several joint works mainly with Hau-Tieng Wu (Courant, NYU).