1: \begin{abstract}

2: This paper studies a class of exponential family models whose

3: canonical parameters are specified as linear functionals of an

4: unknown infinite-dimensional slope function. The optimal minimax

5: rates of convergence for slope function estimation are established.

6: The estimators that achieve the optimal rates are constructed by

7: constrained maximum likelihood estimation with parameters whose

8: dimension grows with sample size. A change-of-measure argument,

9: inspired by Le~Cam's theory of asymptotic equivalence, is used to

10: eliminate the bias caused by the nonlinearity of exponential family

11: models.

12: \end{abstract}