Table 3.
Comparison of akaike information criteria and Bayesian information criteria among different accelerated failure time model and frailty distributions.
| Information criteria | Models | Frailty distributions | ||
|---|---|---|---|---|
| No frailty | Gamma frailty | Inverse Gaussian frailty | ||
| AIC | EXPONENTIAL | 11,996 | 11,998 | 11,998 |
| WEIBULL | 299 . 23 | 966.86 | 808.86 | |
| LOG-LOGISTIC | 946.89 | 944.89 | 944.89 | |
| LOG-NORMAL | 1,054.49 | 1,052.49 | 1,052.49 | |
| GENERALIZED GAMMA | 1,052.69 | 910.81 | 967.50 | |
| BIC | EXPONENTIAL | 12,062.73 | 12,071.4 | 12,071.4 |
| WEIBULL | 225 . 82 | 886.78 | 728.78 | |
| LOG-LOGISTIC | 873.49 | 864.81 | 864.81 | |
| LOG-NORMAL | 981.09 | 972.41 | 972.41 | |
| GENERALIZED GAMMA | 972.61 | 824.06 | 914.12 | |
Bold value indicates the lowest Akaike information criteria and Bayesian information criteria value.