English subtitles

← Outliers Mini-Project Video - Intro to Machine Learning

Get Embed Code
4 Languages

Showing Revision 3 created 05/25/2016 by Udacity Robot.

  1. Welcome to the mini project on outliers.
  2. As you saw in the last lesson,
  3. having large outliers can have a big effect on your regression result.
  4. So in the first part of this mini project,
  5. you're going to implement the algorithm that Sebastian has suggested to us.
  6. So what that means is you take the 10% or
  7. so of data points that have the largest residuals, relative to your regression.
  8. You remove them, and then you refit the regression, and
  9. you see how the result changes.
  10. You'll be implementing that algorithm in this mini project.
  11. The second thing we'll do is take a closer at the Enron data.
  12. This time with a particular eye towards outliers.
  13. You'll find very quickly that there are some data points that fall
  14. far outside of the general pattern.
  15. So we'll talk about these explicitly, and whether this means they should be
  16. removed or they should be given extra special or extra heavy consideration.
  17. It's really cool and I think you will really enjoy it.