Adjust weights for Parameter Estimation
Not urgent. Find a good way to adjust weights for Parameter estimation, such that error is similar.
At the moment error is normalized by mean and number of timepoints. Instead of mean, variance could be more valuable.