Fig. 4.

Convergence of the SMC likelihood estimator with increasing numbers of particles. Mean (top) and coefficient of variation (bottom) of the log-likelihood vs. number of particles used in the our likelihood estimation algorithm. To avoid negative values, the coefficient of variation was calculated by dividing the standard deviation by the absolute value of the mean, and plotted in log₁₀ scale for better visualization. The results were based on 100 repetitions of the likelihood estimation for each particle count