Solving a Sigmoid Function - Wolfram Alpha

  • Context: Undergrad 
  • Thread starter Thread starter inc7
  • Start date Start date
  • Tags Tags
    Function
Click For Summary

Discussion Overview

The discussion revolves around the sigmoid function, specifically its representation and transformation between two forms: \( \frac{1}{1 + e^{-t}} \) and \( 1 - \frac{1}{1 + e^{t}} \). Participants explore how to rewrite the function in this alternate form and discuss the implications of this transformation.

Discussion Character

  • Technical explanation
  • Mathematical reasoning

Main Points Raised

  • One participant seeks help in rewriting the sigmoid function from \( \frac{1}{1 + e^{-t}} \) to \( 1 - \frac{1}{1 + e^{t}} \) and asks for hints or resources.
  • Another participant suggests multiplying the numerator and denominator of \( \frac{1}{1 + e^{-t}} \) by \( e^{t} \) as a potential method for rewriting the function.
  • A later reply provides a series of transformations, showing steps that lead to the alternate form, but does not clarify the reasoning behind the preference for this form.
  • One participant notes that the alternate form illustrates the symmetry of the function, specifically that \( f(t) + f(-t) = 1 \), which may not be immediately apparent from the original expression.

Areas of Agreement / Disagreement

Participants generally agree on the mathematical transformations involved but do not reach a consensus on the reasoning for preferring one form over the other. The discussion remains open regarding the implications of the alternate representation.

Contextual Notes

Some steps in the mathematical transformations are not fully detailed, and the reasoning behind the preference for the alternate form is not explicitly resolved.

inc7
Messages
2
Reaction score
0
Hi,
I'm working with the sigmoid function which is written

1/(1 + e^-t)

While plugging this into wolfram alpha, I noticed an alternate way to write it is

1 - 1/(1 + e^t)

I can't for the life of me figure out how to go about rewriting it in the alternate form. Would anyone be able to give me a hint on how to go about rewriting it this way? Or even better point me at some material that will help me figure it out? Thanks

Here is the link to wolfram alpha

http://www.wolframalpha.com/input/?i=1/(1+++exp(-t))
 
Mathematics news on Phys.org
Start by multipling the top and bottom of 1/(1 + e^-t) by e^t.
 
inc7 said:
Hi,
I'm working with the sigmoid function which is written

1/(1 + e^-t)

While plugging this into wolfram alpha, I noticed an alternate way to write it is

1 - 1/(1 + e^t)

I can't for the life of me figure out how to go about rewriting it in the alternate form. Would anyone be able to give me a hint on how to go about rewriting it this way? Or even better point me at some material that will help me figure it out? Thanks

Here is the link to wolfram alpha

http://www.wolframalpha.com/input/?i=1/(1+++exp(-t))

1/(1 + e-t) = 1/(1 + 1/et) = et/(1 + et) = (1 + et -1)/(1 + et) =

you can complete.

Nothing difficult, only thing difficult to understand maybe is why they want it in that form.
 
epenguin said:
1/(1 + e-t) = 1/(1 + 1/et) = et/(1 + et) = (1 + et -1)/(1 + et) =

you can complete.

Nothing difficult, only thing difficult to understand maybe is why they want it in that form.

Thanks guys. That helped. So to complete would you end up with

1/(1 + et) + 1et/(1 + et) - 1/(1 + et) = 1 - 1/(1 + et)
 
Last edited:
epenguin said:
Nothing difficult, only thing difficult to understand maybe is why they want it in that form.

It shows the function is "symmetrical" in the sense that f(t) + f(-t) = 1, which isn't obvious from either expression on its own.
 

Similar threads

  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 3 ·
Replies
3
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 34 ·
2
Replies
34
Views
3K
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 23 ·
Replies
23
Views
6K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K