Hello everyone!

I know this topic is a little bit far from usual competitive programming, but I was wondering if anyone here could help me in understanding this little concept.

I've been into machine learning lately, and I was wondering about the difference between a "Learning Function" and "Polynomial Interpolation". Consider a dataset where we have a set of points (x0, x1, ..., xn) and their corresponding images (y0, y1, ..., yn). Now if we want to have a neural network that learns the function f(x) = y, isn't that the same as constructing an interpolating polynomial that fits the points? Or is there a hidden difference between the two concepts?