estimate normal distribution parameters by $n$ largest samples

48 Views Asked by At

If I have the $n$ largest out of $m$ values of a sample from independent normal distributed random variables $\mathbb{X}_1,\dots,\mathbb{X}_m\sim\mathcal{N}(\mu,\sigma)$ with unknown parameters $\sigma$ and $\mu$, is there an simple maximum likelihood estimation formula for these?

What for an exponential distribution?