Return to Video

Which Data is Good for PCA - Intro to Machine Learning

  • 0:00 - 0:03
    And the answer is yes in all cases.
  • 0:03 - 0:07
    Now the clear cut case is the left one, which is the one we already discussed.
  • 0:07 - 0:12
    But we have data on a circle that could still be a main axis and
  • 0:12 - 0:14
    a secondary axis.
  • 0:14 - 0:18
    And PCA will actually give you a result, typically related,
  • 0:18 - 0:21
    first X is this direction, second this direction.
  • 0:21 - 0:23
    The third one is surprising.
  • 0:23 - 0:24
    When we, remember regression,
  • 0:24 - 0:27
    it's impossible to build a regression that goes vertically because it
  • 0:27 - 0:33
    can't redivide this data set here as a function f equals f of x.
  • 0:33 - 0:37
    But regression treats the variables very asymmetrically.
  • 0:37 - 0:39
    Minus the input, minus the output.
  • 0:39 - 0:41
    In PCA, all we get is vectors.
  • 0:41 - 0:47
    So I can easily imagine a quadrant system where the x axis falls vertically,
  • 0:47 - 0:51
    and the y axis goes to the left, and that's the answer for PCA in this case.
Title:
Which Data is Good for PCA - Intro to Machine Learning
Description:

more » « less
Video Language:
English
Team:
Udacity
Project:
ud120 - Intro to Machine Learning
Duration:
0:53

English subtitles

Revisions Compare revisions