1: \begin{abstract}

2: Inference for partially observed Markov process models has been a longstanding methodological challenge with many  scientific and engineering applications. 

3: Iterated filtering algorithms maximize the likelihood function for partially observed Markov process models by solving a recursive sequence of  filtering problems. 

4: We present new theoretical results pertaining to the convergence of iterated filtering algorithms implemented via sequential Monte~Carlo filters. 

5: This theory complements the growing body of empirical evidence that iterated filtering algorithms provide an effective inference strategy for scientific models that have been intractable via alternative methodologies.

6: \end{abstract}

7: