1: \begin{abstract}

2: We study the quadratic prediction error method---i.e., nonlinear least squares---for a class of time-varying parametric predictor models satisfying a certain identifiability condition. While this method is known to asymptotically achieve the optimal rate

3: %rate-optimal decay of the prediction error 

4: for a wide range of problems, there have been no \emph{non-asymptotic} results matching these optimal rates outside of a select few, typically linear, model classes.

5: %known to provide parameter estimates that ensure rate-optimal asymptotic convergence of the true prediction error to zero, its finite-sample performance is yet to be characterized, except for a few special---typically linear---model classes. 

6: By leveraging modern tools from learning with dependent data, we provide the first rate-optimal non-asymptotic analysis of this method for our more general setting of nonlinearly parametrized model classes. Moreover, we show that our results can be applied to a particular class of \textit{identifiable} AutoRegressive Moving Average (ARMA) models, resulting in the first optimal non-asymptotic rates for identification of ARMA models. 

7: \end{abstract}

8: