https://communities.sas.com/t5/Statistical-Procedures/Maximum-Likelihood-fitting-of-parameters-that-are-not-predictors/m-p/53926#M2504 Hi All,<BR /> In generalized linear model theory-&gt;maximum likelihood fitting we get the new parameter estimates as<BR /> Beta(r+1) = Beta(r) - (H^-1)*s<BR /> where<BR /> H = -X'WX is the hessian, and<BR /> s =SUM (w(y-Mu)x)/(V(Mu)*g'(Mu)*Phi)) is the gradient vector<BR /> <BR /> My question is how do these formulas change for a parameter that is not Beta. For example the dispersion parameter in negative binomial or Scale parameter in normal, so i can estimate it as an entry in the hessian and gradient vector.<BR /> I would appreciate any reference to some literature or help with the formulas. Thanks! Wed, 20 Apr 2011 19:09:35 GMT sramalingam 2011-04-20T19:09:35Z https://communities.sas.com/t5/Statistical-Procedures/Maximum-Likelihood-fitting-of-parameters-that-are-not-predictors/m-p/53926#M2504 Hi All,<BR /> In generalized linear model theory-&gt;maximum likelihood fitting we get the new parameter estimates as<BR /> Beta(r+1) = Beta(r) - (H^-1)*s<BR /> where<BR /> H = -X'WX is the hessian, and<BR /> s =SUM (w(y-Mu)x)/(V(Mu)*g'(Mu)*Phi)) is the gradient vector<BR /> <BR /> My question is how do these formulas change for a parameter that is not Beta. For example the dispersion parameter in negative binomial or Scale parameter in normal, so i can estimate it as an entry in the hessian and gradient vector.<BR /> I would appreciate any reference to some literature or help with the formulas. Thanks! Wed, 20 Apr 2011 19:09:35 GMT https://communities.sas.com/t5/Statistical-Procedures/Maximum-Likelihood-fitting-of-parameters-that-are-not-predictors/m-p/53926#M2504 sramalingam 2011-04-20T19:09:35Z