“…Approaches for dealing with this issue include, for example, approximating the likelihood using Gaussian or linear approximations (Julier and Uhlmann, 1997;Wan and Van Der Merwe, 2000), numerical approximation using Laplace or HMM approximations (Bucy and Senne, 1971;Koyama et al, 2010;Thygesen et al, 2017), the use of sequential Monte Carlo to obtain an unbiased approximation to the likelihood (Andrieu and Roberts, 2009), and data augmentation using sequential Monte Carlo, numerical approximation, and/or MCMC approaches (Frühwirth-Schnatter, 1994;Lindsten et al, 2014;Fearnhead and Meligkotsidou, 2016). A review of these methods, and others, is given in Newman et al (2022).…”