Hi I am wondering if there is a way to linearize y=exp(x) where x is a continuous variable? I know Taylor series would give a polynomial approximation which is not linear

Thanks

asked 01 Mar '13, 15:50

hesameivazy's gravatar image

hesameivazy
712
accept rate: 0%


In general, there is no exact linearization for exp(x), or any general continuous nonlinear function for that matter.

However, you can approximate it by:

  1. Using piecewise-linear segments in the domain of interest (this can be effective if the domain of x is small)
  2. A secant line, if the domain of interest is really small, and you are wiling to tolerate a reasonable amount of error.

If the above options are inappropriate for your problem, you may want to consider either finding an alternative formulation that does not call for the use of exp(x), or solve your problem as a nonlinear program (if it's a purely continuous problem).

If it contains integer variables, you may want to solve it as an MINLP (though this is rarely advisable except for fairly small problems with very few binary variables).

link

answered 01 Mar '13, 17:59

Gilead's gravatar image

Gilead ♦
2.0k39
accept rate: 17%

edited 01 Mar '13, 18:07

Your answer
toggle preview

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here

By RSS:

Answers

Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text](http://url.com/ "Title")
  • image?![alt text](/path/img.jpg "Title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported

Tags:

×96
×86
×40

Asked: 01 Mar '13, 15:50

Seen: 641 times

Last updated: 01 Mar '13, 18:07

OR-Exchange! Your site for questions, answers, and announcements about operations research.