Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Suppose we have 4 normally distributed variables X1, X2, X3 and X4. The covariance matrix and mean values of X1 and X2 are given. Is there a way to determine the mean values of variables X3 and X4?

share|improve this question
    
No, there is no way to estimate the unknown means in general. –  Dilip Sarwate Apr 7 '12 at 23:11
1  
The covariances are based on the centered random variables X-E(X), hence they cannot feel the expectations E(X). –  Did Apr 7 '12 at 23:21
    
I forgot to mention the fact that the variables are all normally distributed, but I guess that doesn't change much. –  Jacques Apr 8 '12 at 6:06

1 Answer 1

No, you know nothing about the means of $X_3$ and $X_4$.

Suppose you let $Y_3 = X_3 + k_3$ and $Y_4 = X_4 + k_4$ for real constants $k_3$ and $k_4$.

Then looking at $X_1$, $X_2$, $Y_3$ and $Y_4$: the means of $X_1$ and $X_2$ would still be the same, and the covariance matrix would also be the same, even though the means of the third and fourth variables had changed by arbitrary amounts.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.