I found sligthly different results elsewhere !

Curve fitting equations are used to fit your data to a theoritical curve. Several methods exist to do this job.
One method, first transforms your data into a line, then applies to this line the least squares fitting method. Another method (the one used here), applies least square fitting method directly to your data. This method was used because it weights the data points equally. You will find some explanations with an example plot here.

Weisstein, Eric W. Least Squares Fitting--Exponential.From MathWorld--A Wolfram Web Resource.
http://mathworld.wolfram.com/LeastSquaresFittingExponential.html.