I need help

b. Let Y = In(Z) with Y ~ N(#, o). Suppose the distribution of a predictor of a new value of Y follows a normal distribution, i.e. Yo ~ N(u, va'Sa) where a is a vector of constants and > is the variance covariance matrix of Y. (We assume that Yo = a'Y, a linear combination of Y = (Y1, Y2, ..., Ym)'.) In addition suppose Yo = In(Zo). Find the unbiased predictor of Zo