Table 7.

Model selection for the $\mathrm{\lambda }$ RM promoter data. The columns show the maximum likelihood estimates, the loglikelihood values, and the AICs, respectively

		${\mathit{q}}_{\mathbf{on}}$	${\mathit{q}}_{\mathbf{off}}$	${\mathbf{\lambda }}_1$	${\mathbf{\lambda }}_2$	${\mathbf{\lambda }}_3$	$\mathbf{log}\mathcal{L}$	AIC
Seq-1	–	–	–	0.0144	–	–	− 3569.9	7141.8
Seq-2	${\mathrm{\lambda }}_1\le {\mathrm{\lambda }}_2$	–	–	0.0144	8.9245	–	− 3569.4	7142.7
Seq-2	${\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_1$	–	–	5.8876	0.0144	–	− 3569.5	7143.1
Seq-3	${\mathrm{\lambda }}_1\le {\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_3$	–	–	0.0144	9.9536	9.9875	− 3569.3	7144.6
Seq-3	${\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_1\le {\mathrm{\lambda }}_3$	–	–	5.8876	0.0144	6.0453	− 3570.5	7147.0
Seq-3	${\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_3\le {\mathrm{\lambda }}_1$	–	–	5.8876	0.0144	5.7305	− 3570.7	7147.3
On/off-seq-1	–	0.0249	0.0608	0.0496	–	–	− 3475.2	6956.5
On/off-seq-2	${\mathrm{\lambda }}_1\le {\mathrm{\lambda }}_2$	0.0303	0.4089	0.2220	0.2221	–	− 3474.8	6957.6
On/off-seq-2	${\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_1$	0.0254	1.2284	0.7920	0.1416	–	− 3468.2	6944.4
On/off-seq-3	${\mathrm{\lambda }}_1\le {\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_3$	0.0312	0.4410	0.2314	0.2314	9.9983	− 3475.6	6961.3
On/off-seq-3	${\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_1\le {\mathrm{\lambda }}_3$	0.0255	2.4371	1.5558	0.1427	3.8107	− 3466.5	6943.1
On/off-seq-3	${\mathrm{\lambda }}_2\le {\mathrm{\lambda }}_3\le {\mathrm{\lambda }}_1$	0.0254	4.4154	2.8221	0.1423	2.6958	− 3466.9	6943.7