Let $X_1, X_2, \dots, X_n$ be i.i.d. random variables with PDF $f(x; \theta) = \frac{1}{\theta} x^{(1-\theta)/\theta}$ for $0 < x < 1$ and $\theta > 0$. What is the Maximum Likelihood Estimator (MLE) for $\theta$?