Fig. 1: Identification of coefficients in the simulated personal temporal symptom networks.
From: Estimating personal temporal symptom networks for childhood cancer survivors
The evaluation metrics were assessed with different sample sizes. TPR: true positive rate. FPR: false positive rate. MCC: Matthew’s correlation coefficient. \(\eta\): the tuning parameter of eBIC that controls the network sparsity. The dots represent the average values, and the error bars denote 95% CIs.
