Determining Acceleration Based On Graph I understand how to solve this problem, but I am unsure how to generate an equation for the graph (below).  My current attempt involves using the mass provided along with the derivative of the line (acceleration) to calculate the force (Newtons Second Law). 
Usually, point slope form would work. However, it's a curve, so the only thing point-slope can do is give me an approximation.  I feel like I am missing something. Maybe the approximation is the right way to go?
 A: How you approach this depends on the time and resources you have available to you.
If a problem like this cropped up as part of a scientist's work the best approach would be to try and understand the underlying physics. You would develop a mathematical model for the system then fit the data to a function derived from that model. In this case the model would be complicated as you have quadratic aerodynamic drag and mechanical drag that is probably (but not necessarily) linear in the velocity.
The next best option is to just choose some function based on how well you think it would fit rather than because it has any physical significance. In this case 5 minutes with copy of Excel produced a fit:

with the pink line showing the fitted function:
$$ v = v_0 e^{-0.0162 t} $$
where $v_0$ is the speed at $t = 0$, and the speed has been converted to metres per second. You can differentiate the fitted function to get a fitted acceleration and therefore the force. NB I had to read the points off your graph so there will be an additional source of error due to this.
If the question cropped up in an exam I'd guess the intention is that you just measure the tangent by eye using a ruler.
A: At each point you have $F=ma$.  The instantaneous acceleration is the slope of the speed-time curve.  It looks like acceleration may be a linear function of time here. You could pick three points and calculate the acceleration and speed as a function of time and see if they match the curve.  It will be hard, as errors will depend on the error of the speed from your model.  The other choice is to use a ruler to approximate the tangent to the curve, get the slope, consider that the acceleration, and again use $F=ma$
