all:
I wonder what are the differences between optimization and filtering. E.g. Newton-Gauss method vs Kalman Filter. What are their pros and cons? Also, could you illustrate the differences with some practical problems, such as which approach is more suitable for some sort of problems. Many thanks in advance.
Regards
The Kalman filter is an explicit analytical solution to the special case that you minimize the variance of the estimation errors under assumptions on normally distributed process and measurement noise, and linear system dynamics. Hence, if that is the problem you have, no reason apply numerical optimization.