壓縮感知

Compressed sensing is regarded as a promising solution for data com- pression and analytics which has attracted great attention recently. This technique dramatically reduces the volume of original data and has high probability to recover the original data with less information loss. However, it is applicable only when the input data has the sparse property, which is not the case for most real world data. Constructing a sparse representation for the raw data becomes the most important step before we apply compressed sensing to tackle the problems. In this course, we will focus on the fundamen- tal theory for compressed sensing, optimization techniques for the recovery algorithms and the applications of compressed sensing. My ultima goal is exploring the possibility to run machine learning tasks in the compressed domain directly.

Linear Algebra Probability Mathematical analysis Numerical Methods

無備註

教師未提供此項資料

Homework: 40% Final Exam: 30% Presentation: 30%

教師未提供此項資料

M. A. Davenport, M. F. Duarte, Y. C. Eldar, and G. Kutyniok. Introduction to compressed sensing. In Compressed Sensing: Theory and Applications. Cambridge University Press, 2012

地點

SA 240

時間

09:00~10:10 on Wednesday or e-mail to make an appointment

聯絡方式

yuhjye@math.nctu.edu.tw