English subtitles

← Linearly Separable Data - Intro to Machine Learning

Get Embed Code
5 Languages

Showing Revision 4 created 02/21/2017 by Shuji Watanabe.

  1. So one method in machine learning that's very,
  2. very popular is called decision trees.
  3. And just as in support vector machines, you were able to use a kernel trick,
  4. to change from linear to non-linear decision surfaces.
  5. Decision trees use a trick to let you do non-linear decision making with simple,
  6. linear decision surfaces.
  7. So let's start with an example.
  8. I have a friend named Tom, and Tom loves to windsurf.
  9. But to enjoy wind surfing, he needs two things.
  10. He needs wind, and he also needs the sun because he
  11. doesn't really like surfing with overcast or in rain.
  12. So, we take all the days of last year and make them data points.
  13. He doesn't surf when the weather isn't very sunny.
  14. And he doesn't surf when there's not enough wind.
  15. But when these conditions are met as windy, and it's sunny,
  16. he enjoys windsurfing on the lake.
  17. My very first question, is this data linearly separable?
  18. Yes or no?