ncku-talk2018-01-11

Colloquium

DATE 2018-01-11　16:10-17:00

PLACE 數學館3174教室

SPEAKER 陳志偉博士（國家理論科學中心數學組）

TITLE Geometric Analysis and Data Representation

ABSTRACT One of the main issue in manifold learning theory is to represent the data set by appropriate embedding maps. Such process is expected to reduce the dimension of the data set and maintain as much geometric structures as possible. One method is to employ eigenfunctions of the Laplacian as embedding functions. Surprisingly, this technique could be applied to tackle problems occurring in an active area in modern geometric analysis - the Ricci flow. In this talk, I will show you the confluence of the Ricci flow and the manifold learning theory.

Colloquium

DATE	2018-01-11　16:10-17:00

PLACE	數學館3174教室

SPEAKER	陳志偉博士（國家理論科學中心數學組）

TITLE	Geometric Analysis and Data Representation

ABSTRACT	One of the main issue in manifold learning theory is to represent the data set by appropriate embedding maps. Such process is expected to reduce the dimension of the data set and maintain as much geometric structures as possible. One method is to employ eigenfunctions of the Laplacian as embedding functions. Surprisingly, this technique could be applied to tackle problems occurring in an active area in modern geometric analysis - the Ricci flow. In this talk, I will show you the confluence of the Ricci flow and the manifold learning theory.