difference between estimation and its mean

28 Views Asked by At

I am calculating covariance between estimated means (gaussian), $Cov( \hat x_{t-1}, \hat y_{t-1}) = \mathbb{E}[(\hat x_{t-1} - \mu_{x,t-1}) (\hat y_{t-1} - \mu_{y,t-1})] = ?$ As i understand, the mean of the posterior at t-1 is estimation at t-1, is the difference then =0? I believe this shouldn't be correct, but what is it then?