Return to Video

02-38 Numerics

  • 0:00 - 0:04
    And this course we are mostly using the computer to do simulations
  • 0:04 - 0:07
    numerically, that is by number crunching.
  • 0:07 - 0:09
    The major reason for that is that most problems can't be solved
  • 0:09 - 0:12
    by clean equations, unlike school book problems.
  • 0:12 - 0:15
    The real road is way too complex to do that.
  • 0:15 - 0:19
    A second reason is that particularly we have to base our simulations
  • 0:19 - 0:27
    on huge numbers of measurements be it climate's data, be it financial data, be it data from biology,
  • 0:27 - 0:30
    but the numerically approach also has a series of drawbacks.
  • 0:30 - 0:35
    In particular, it's very hard to prove general results using numerics.
  • 0:35 - 0:40
    Is there no way that a particle in some system escapes to infinity?
  • 0:40 - 0:47
    You can look at a single trajectory at a time, but you cannot look at an infinite number of trajectories
  • 0:47 - 0:51
    and numerical methods almost always introduce discretization errors
  • 0:51 - 0:53
    also known as truncation errors.
  • 0:53 -
    In particular, we have to work with finite steps in time.
Title:
02-38 Numerics
Team:
Udacity
Project:
CS222 - Differential Equations
Duration:
0:59
Amara Bot edited English subtitles for 02-38 Numerics
Gundega added a translation

English subtitles

Revisions Compare revisions