埃马纽埃尔-坎代斯 数学家和统计学家

shiyi18

Emmanuel Candès
Mathematician and Statistician | Class of 2017
Exploring the limits of signal recovery and matrix completion from incomplete data sets with implications for high-impact applications in multiple fields.


Portrait of Emmanuel Candès
Photos for download >
Title
Mathematician and Statistician
Affiliation
Department of Statistics, Stanford University
Location
Stanford, California
Age
47 at time of award
Area of Focus
Mathematics, Statistics, and Probability, Computer Science and Electrical Engineering
Website
Stanford University: Emmanuel Candès
Published October 11, 2017
ABOUT EMMANUEL'S WORK
Emmanuel Candès is a mathematician and statistician known for developing a unified framework for addressing a range of problems in engineering and computer science, most notably compressed sensing. Compressed sensing is a technique for efficiently reconstructing or acquiring signals that make up sounds and images. Candès's research focuses on reconstructing high-resolution images from small numbers of random measurements, as well as recovering the missing entries in massive data tables.

Using an approach that draws on concepts from linear algebra and L1 minimization (a concept of high-dimensional geometry), Candès and colleagues were able to reconstruct high-resolution signals from sparse measurements under specified conditions. In diagnostic healthcare, for example, reducing the number of measurements needed to create high-resolution MRI scans shortens the amount of time patients must remain still in the scanner, an outcome with particularly beneficial implications for children. The ability to process and/or reconstruct audio, visual, and wireless signals from limited data has also led to significant refinements in digital photography, radar imaging, and wireless communications. Candès has expanded this work to address problems in low-rank matrix completion, devising statistical estimation methods for inferring missing entries in data arrays. (This is analogous to trying to identify a customer's movie preferences from the partial movie ratings that the user has provided.) His framework holds promise for phase retrieval, a problem arising in many applications such as crystallography, diffraction imaging (X-ray), and astronomical instrumentation.

Candès's work at the interface of applied and theoretical mathematics is generating new lines of research in information theory as well as laying the groundwork for improvements in many devices that make use of signal and image processing methods.

BIOGRAPHY
Emmanuel Candès received a B.E. (1993) from École Polytechnique, an M.Sc. (1994) from Université de Paris VI, and a Ph.D. (1998) from Stanford University. He was a member of the faculty of Stanford University (1998–2000) and the Department of Computing and Mathematical Sciences at the California Institute of Technology (2000–2009), before returning to Stanford as the Barnum-Simons Chair in Mathematics and Statistics in the Departments of Mathematics and Statistics and a professor of electrical engineering (by courtesy). His scientific papers have been published in IEEE Transactions on Information Theory, Annals of Statistics, Communications on Pure and Applied Mathematics, IEEE Signal Processing Magazine, and Proceedings of the National Academy of Sciences.



埃马纽埃尔-坎代斯
数学家和统计学家 | 2017级
探索不完整数据集的信号恢复和矩阵完成的极限，对多个领域的高影响应用产生影响。


埃马纽埃尔-坎戴斯的画像
图片下载 >
标题
数学家和统计学家
工作单位
斯坦福大学统计系
工作地点
斯坦福, 加州
年龄
获奖时47岁
重点领域
数学、统计和概率，计算机科学和电气工程
网址
斯坦福大学。Emmanuel Candès
发表于2017年10月11日
关于Emmanuel的工作
Emmanuel Candès是一位数学家和统计学家，以开发一个统一的框架来解决工程和计算机科学中的一系列问题而闻名，其中最引人注目的是压缩传感。压缩传感是一种有效重建或获取构成声音和图像的信号的技术。Candès的研究重点是通过少量的随机测量来重建高分辨率的图像，以及恢复大量数据表中的缺失项。

使用一种借鉴线性代数和L1最小化（高维几何学的一个概念）的方法，Candès及其同事能够在特定条件下从稀疏的测量中重建高分辨率的信号。例如，在医疗诊断中，减少创建高分辨率核磁共振扫描所需的测量数量，可以缩短病人在扫描仪中必须保持静止的时间，这一结果对儿童特别有利。从有限的数据中处理和/或重建音频、视觉和无线信号的能力，也导致了数字摄影、雷达成像和无线通信的重大改进。Candès将这项工作扩展到解决低等级矩阵完成的问题，设计了统计估计方法来推断数据阵列中的缺失条目。(这类似于试图从用户提供的部分电影评级中识别客户的电影偏好）。他的框架有望用于相位检索，这是一个在许多应用中出现的问题，如晶体学、衍射成像（X射线）和天文仪器。

Candès在应用数学和理论数学界面的工作正在产生信息理论的新研究方向，并为许多利用信号和图像处理方法的设备的改进奠定了基础。

个人简历
Emmanuel Candès于1993年获得巴黎综合理工学院的工程学学士学位，1994年获得巴黎第六大学的理学硕士学位，1998年获得斯坦福大学的博士学位。他曾是斯坦福大学（1998-2000年）和加州理工学院计算机和数学科学系（2000-2009年）的教师，之后回到斯坦福大学担任数学和统计系的巴纳姆-西蒙斯讲座教授和电气工程教授（礼聘）。他的科学论文发表在IEEE Transactions on Information Theory, Annals of Statistics, Communications on Pure and Applied Mathematics, IEEE Signal Processing Magazine, and Proceedings of the National Academy of Sciences。