Did a lot of Google searching, but couldn't find a common-sense answer. For non linear equations like y = a + b * x/2 + c * ln(x^2), I can see how substituting various values on the right hand side of the equation can yield a curve defined by the equation.
What I don't understand is how the equation was created. If I have some data points on a graph that appear to follow the pattern of an exponent, I could use a function like y = e^x to describe the data.
But if I have data points on a graph that appear like a more strange looking curve for which I don't know what equation might best describe it, how is it possible to create an equation that describes the curve? I'm puzzled at how somebody came up with the x/2 and ln(x^2) (above equation) to describe some curve.