MAXIMUM LIKELIHOOD ESTIMATION IN GENERALIZED GAMMA TYPE MODEL
In the present paper, the maximum likelihood estimates of the two parameters of a generalized gamma type model have been obtained directly by solving the likelihood equations as well as by reparametrizing the model first and then solving the likelihood equations (as done by Prentice, 1974) for fixed values of the third parameter. It is found that reparametrization does neither reduce the bulk nor the complexity of calculations. as claimed by Prentice (1974). The procedure has been illustrated with the help of an example. The distribution of MLE of q along with its properties has also been obtained.