|Manifold denoising under high dimensional noise with clinical applications
|Hau-Tieng Wu氏 (Courant Inst., NYU)
|The analysis of noisy and high-dimensional clinical data stands as a pressing challenge in medical research. "Manifold denoising" is a data sharpening technique that unravels meaningful patterns while mitigating noise under the manifold model.
I will discuss recent progress made in this direction, emphasizing the pivotal role played by random matrix theory and spectral geometry as our primary analytical tools.
We will not only delve into the theoretical underpinnings but also demonstrate one clinical example and its results, all purposefully structured to offer pragmatic solutions in the quest to refine healthcare data analysis.