I have an exponential regression equation that I use to predict the future condition of roads:
$$y=a+be^{cx}$$
Modified for my purposes as
$$y=21-e^{a x}$$
Using the normal equation
$$a=\frac{\sum_{i=1}^n x_iz_i } { \sum_{i=1}^n x_i^2 }=\frac{\sum_{i=1}^n x_i \log(21-y_i)} { \sum_{i=1}^n x_i^2 }$$
I've come up with an initial estimate for $a$
C9 =21-EXP($C$33*A9)
C33 {=SUM(VALUE(A9:A29*LN(21-B9:B29))) / SUM(VALUE(A9:A29^2))}
C34 {=SUM(VALUE( (B9:B29 - C9:C29)^2 ))}
Question:
I want to tune the estimate using the Newton method (as described here ) to minimize the ESS. How can I do that?

With the help of a colleague, I was able to tune the estimate using three Newton-Raphson iterations:
The (n+1)th iteration is given by:
$$a_{n+1}=a_n-\frac{f(a_n)}{f'(a_n)}$$
Initial approximation $_0$ = 0.147233112 (coefficient found through Normal Equations)
$$f(a)=\sum_{i=1}^n \left(21-e^{ax_i}-y_i \right)e^{ax_i}x_i$$
$$f'(a)=\sum_{i=1}^n \left(21-2e^{ax_i}-y_i \right)e^{ax_i}x_i^2$$
We stop when further iterations don't lead to a further reduction in the ESS. In this case, only three iterations were needed.
Excel 2016:
Formulas:
Notes:
As mentioned, I can't take credit for the calculus stuff. A colleague helped me with that.
I'd be happy to hear about any mistakes or possible improvements. Layman's terms would be appreciated.