Let x = (x1, . . . , xn1) and censored observations (xn1 +1, . . . ,xn) (that is, ith experiment, if I > n1, the survival time is at least yi). Let the new complete censored data yi be such that

Let the mean survival time be θ and the probability density of y be

and let the survival function be defined as the probability that an individual survives beyond time y, that is, S(y) = P(Y > y). Thus,

(a) Obtain the MLE, θ-capM:
(b) Obtain an EM algorithm.

Transcribed Image Text:

x, ni