2
$\begingroup$

When speaking of neural networks, I don't get the difference between nonlinear and non-deterministic. Basically, both say that the output of something is not directly correlated to the input?

Hope someone can illuminate me.

$\endgroup$
2
$\begingroup$

Non linear means that there may be a correlation with the input, but it is not a linear one. A function is say to be linear in some argument (the input) when the ratio result/argument is constant. In your case, it is the ratio output/input. Actually, the mathematical definition is a bit more general.

Non determinism is of a different nature. The input/output relation is said to be non-deterministic when one of several result may occur, without any a priori known cause. This is usually modelelled mathematically, either by using a relation rather than a function, of by considering a function from the input domain to the domain of subsets of the output domain.

For example: if you consider inputs and outputs in the domain of integer, for each integer input, you have a set of possible outputs that are all integers.

When this set always contains only a single element, the function is deterministic, and the set can be replaced by this unique element.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.