Continuous Optimization Seminar- Courtney PaquetteExport this event to calendar

Thursday, March 7, 2019 — 4:00 PM EST

Title: Introduction to high-dimensional probability: some basic concentration inequalities and useful distributions

Speaker: Courtney Paquette
Affiliation: University of Waterloo
Room: MC 5417

Abstract: In this seminar, we introduce important tools from high-dimensional probability useful in studying applications in data science such as covariance estimation, matrix completion, elements of statistical learning, networks, semidefinite programming, and many more. High-dimensional probability is an area of probability theory that studies random objects in R^n where the dimension of n is large. We will learn, in the seminar, basic theoretical skills (e.g. concentration inequalities, sub-gaussian distributions) for analyzing random vectors, random matrices, and random projections.  An outline of the seminar is as follows: (1) we will cover Chapters 2-6 from “High-dimensional probability” by Roman Vershynin (available online) and (2) we will present papers in data science which utilize these tools.

In this talk, I introduce everyone to the rich topic of concentration inequalities – inequalities that quantify how a random variable deviates around its mean. In particular, I will prove some basic inequalities: Hoeffding, Chernoff, and Bernstein. Then I will present two important classes of distributions: sub-gaussian and sub-exponential. These classes form a natural environment in which many results of high-dimensional probability and its applications are developed from. 

Location 
MC - Mathematics & Computer Building
5417
200 University Avenue West

Waterloo, ON N2L 3G1
Canada

S M T W T F S
29
30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1
2
  1. 2019 (141)
    1. October (12)
    2. September (15)
    3. August (9)
    4. July (17)
    5. June (18)
    6. May (16)
    7. April (9)
    8. March (24)
    9. February (13)
    10. January (8)
  2. 2018 (138)
    1. December (2)
    2. November (18)
    3. October (14)
    4. September (9)
    5. August (2)
    6. July (10)
    7. June (13)
    8. May (17)
    9. April (9)
    10. March (19)
    11. February (14)
    12. January (11)
  3. 2017 (103)
  4. 2016 (137)
  5. 2015 (136)
  6. 2014 (88)
  7. 2013 (48)
  8. 2012 (39)
  9. 2011 (36)
  10. 2010 (40)
  11. 2009 (40)
  12. 2008 (39)
  13. 2007 (15)