Hi all, Well, most neurons work with inputs that handle 1/0. Are there any neurons designed to handle a certain range of signal amplitude? like, an input that can go from 0 to 100? How does the neuron handle it? what equations are envolved? also, doing this one must realize that the neurons will also have to output values in the range of 0 to 100. How to do this? the firing equation would have to be diferent... some input on this plz...

[Hugo Ferreira][Positronic Dreams][]
"Somewhere, something incredible is waiting to be known."
- Carl Edward Sagan (1934 - 1996)

Most neural network implementations I''ve seen handle numbers from 0 to 1, inclusive and not just the two integers. If you had such an implementation, it would be easy to make it handle input/output in the range of 0 to 100. You just need to divide the input by 100 before you feed it into the network, and multiply the output by 100 to put it in the desired range. Read fup''s tutorial on his website at http://www.btinternet.com/~fup/Stimulate.html for info on implementing a simple network that handle values in the range 0 to 1.

"The Requested Information Is Unknown Or Classified" -Anonymous

To add:

A neuron may have any input range. There is no minimum or maximum. It is advisable to *standardize* all your inputs though.

I assume when you say you need your neurons to output in the range 0-100 that what you meant to say is that you want the *network* to output in that range. As Extrarius says all you have to do is multiply the output by the correct figure.

By the sounds of it you need to learn a lot more about *activation functions*. When you understand them a little more your path will become clear.

Hope this helps

---

Stimulate