Referee report for "Efficient Bayesian inference for large chaotic dynamical systems"
2020
This paper aims to solve the high computational challenge in the statistical inference of chaotic dynamic models.This is a practical and challenging problem.It has great potential in practice.This paper combines two methods to solve the intractability of inference of chaotic dynamic models.CIL was adopted to incorporate more information of the observations into the summary statistics.LA-MCMC was utilized to reduce computational time and thus make the proposed method more practical.The paper was written in a very good manner.First of all, the problem was introduced in Section 1. Then CIL and LA-MCMC were described in detail in Section 2. Three examples were demonstrated in Section 3. The authors have put a lot of effort to CIL and LA-MCMC, especially CIL.This makes the paper self-contained.To improve the C1