Department Seminar by Paul Garthwaite, the Open UniversityExport this event to calendar

Thursday, May 25, 2017 — 4:00 PM EDT

‘Tweaking’ variables to make them uncorrelated

Interpreting the results of a statistical analysis would be easier if all variables were uncorrelated and all data sets consisted of orthogonal x-variables. Principal components is one way of obtaining orthogonal/uncorrelated variables, but the relationship between the principal components and the original x-variables is complex. This talk proposes two transformations that give orthogonal/uncorrelated components with a close one-to-one correspondence between the x-variables and the components.

The transformations have a number of useful applications including:

  • a unified approach to the identification and diagnosis of collinearities
  • evaluating the contributions of individual variables to a Mahalanobis distance or the test statistic of a Hotelling test
  • setting prior weights for Bayesian model averaging
  • evaluating the contributions of individual variables to a multiple regression
  • calculation of an upper bound for a multivariate Chebyshev inequality.

The talk will focus on the first three of these applications.

Location 
M3 - Mathematics 3
Room: 3127
200 University Avenue West
Waterloo, ON N2L 3G1
Canada

S M T W T F S
25
26
27
28
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
3
4
5
  1. 2017 (33)
    1. December (1)
    2. October (1)
    3. September (3)
    4. July (1)
    5. June (2)
    6. May (4)
    7. April (2)
    8. March (3)
    9. February (4)
    10. January (12)
  2. 2016 (44)
    1. December (2)
    2. November (4)
    3. October (4)
    4. September (5)
    5. August (2)
    6. June (5)
    7. May (3)
    8. April (1)
    9. March (5)
    10. February (3)
    11. January (10)
  3. 2015 (38)
  4. 2014 (44)
  5. 2013 (46)
  6. 2012 (44)