RLMD-PA: A Reinforcement Learning-Based Myocarditis Diagnosis Combined with a Population-Based Algorithm for Pretraining Weights

Table 7

Results of 5-CV classification performances (accuracy, recall, and precision) obtained for automated myocarditis detection using various conventional and metaheuristic algorithms with the Z-Alizadeh Sani myocarditis dataset.

	Accuracy					Recall					Precision
Method	Min	Median	Max	Mean	Std.dev.	Min	Median	Max	Mean	Std.dev.	Min	Median	Max	Mean	Std.dev.
CNN + GDM + RL	0.811	0.857	0.868	0.849	0.022	0.784	0.801	0.830	0.806	0.018	0.732	0.806	0.825	0.796	0.038
CNN + GDA + RL	0.817	0.846	0.857	0.840	0.017	0.784	0.812	0.837	0.808	0.022	0.742	0.786	0.828	0.778	0.035
CNN + GDMA + RL	0.829	0.855	0.887	0.854	0.025	0.764	0.816	0.855	0.817	0.037	0.752	0.809	0.849	0.800	0.037
CNN + OSS + RL	0.823	0.849	0.867	0.846	0.016	0.741	0.814	0.837	0.804	0.037	0.778	0.787	0.814	0.791	0.015
CNN + BR + RL	0.826	0.833	0.855	0.837	0.012	0.745	0.796	0.812	0.785	0.027	0.752	0.761	0.850	0.784	0.041
CNN + GWO + RL	0.833	0.848	0.869	0.850	0.016	0.771	0.796	0.842	0.804	0.027	0.769	0.800	0.816	0.797	0.020
CNN + BAT + RL	0.837	0.847	0.865	0.851	0.013	0.778	0.782	0.833	0.796	0.024	0.787	0.805	0.830	0.807	0.016
CNN + COA + RL	0.815	0.843	0.882	0.844	0.028	0.750	0.826	0.856	0.813	0.046	0.748	0.757	0.838	0.781	0.039
CNN + WOA + RL	0.820	0.845	0.847	0.837	0.012	0.750	0.826	0.814	0.789	0.021	0.742	0.783	0.807	0.781	0.024