By the factorization theorem the largest order stat, i.e., the maximum of your data set is sufficient for i.
Next find it's density and from that it's expectation.
Let T be a sample of size s randomly chosen from independent values from a uniform distribution F=F(y,x) on the interval [0,x], and x is not known to us.
is estimator i=T(max) will always be sufficient for i?
is there any way that we can find a unbiased estimator for it
if not how do we prove it?
Thank you