fareb
Newbie level 5
Hi
I need to obtain variance-covariance matrix by maximum likelihood. i obtained hessian matrix and now I need to calculate expected value of this matrix, but I have a lot of problem with calculating expected value. for example I do not understand why in document it is mentioned E[X'X]=X'X or E[X'(Y-XB)]=0
XB is my model that was fitted to my data and Y is my observation. I consider a normal distribution for my data.
I have another models too. Model2=B1*integral(X)+B0*X
what I should do for these models and variance matrix, especially I have problem in calculating expected value.
I would be grateful if any one can help me or guide me to learn more about this.
Thanks
Best wishes
I need to obtain variance-covariance matrix by maximum likelihood. i obtained hessian matrix and now I need to calculate expected value of this matrix, but I have a lot of problem with calculating expected value. for example I do not understand why in document it is mentioned E[X'X]=X'X or E[X'(Y-XB)]=0
XB is my model that was fitted to my data and Y is my observation. I consider a normal distribution for my data.
I have another models too. Model2=B1*integral(X)+B0*X
what I should do for these models and variance matrix, especially I have problem in calculating expected value.
I would be grateful if any one can help me or guide me to learn more about this.
Thanks
Best wishes