Skip to main content
. 2023 Jul 25;28(10):4225–4233. doi: 10.1038/s41380-023-02174-0

Table 1.

Fit statistics from the mixture model.

Model AIC BIC sBIC Entropy Post. Prob.
2-class 8704028.27 8704661.99 8704477.67 0.957 0.986
3-class 8587074.65 8588003.38 8587733.25 0.921 0.869
4-class 7567653.29 7568877.03 7568521.09 0.980 0.989
5-class 7519572.55 7521091.30 7520649.55 0.940 0.895
6-class 7478289.65 7480103.41 7479575.85 0.983 0.939
7-class 7437821.89 7439930.65 7439317.29 0.952 0.885
8-class 7268011.70 7270415.47 7269716.30 0.908 0.826

The bolded row represents the chosen solution.

AIC Akaike’s information criteria, BIC Bayesian information criteria, sBIC sample-size-adjusted Bayesian information criteria, Post. Prob. Average posterior probabilities of membership in assigned class.